Pain

More people with fever and body aches are turning to NSAIDs to ease symptoms, but the drugs have come under new scrutiny as investigators work to determine whether they are a safe way to relieve the pain of COVID-19 vaccination or symptoms of the disease.

Early on in the pandemic, French health officials warned that NSAIDs, such as ibuprofen, could worsen coronavirus disease, and they recommended switching to acetaminophen instead.

The National Health Service in the United Kingdom followed with a similar recommendation for acetaminophen.

But the European Medicines Agency took a different approach, reporting “no scientific evidence” that NSAIDs could worsen COVID-19. The U.S. Food and Drug Administration also opted not to take a stance.

The debate prompted discussion on social media, with various reactions from around the world. It also inspired Craig Wilen, MD, PhD, from Yale University, New Haven, Conn., and associates to examine the effect of NSAIDs on COVID-19 infection and immune response. Their findings were published online Jan.20 in the Journal of Virology.

“It really bothered me that non–evidence-based decisions were driving the conversation,” Dr. Wilen said. “Millions of people are taking NSAIDs every day and clinical decisions about their care shouldn’t be made on a hypothesis.”

One theory is that NSAIDs alter susceptibility to infection by modifying ACE2. The drugs might also change the cell entry receptor for SARS-CoV-2, alter virus replication, or even modify the immune response.

British researchers, also questioning the safety of NSAIDs in patients with COVID-19, delved into National Health Service records to study two large groups of patients, some of whom were taking the pain relievers.

“We were watching the controversy and the lack of evidence and wanted to contribute,” lead investigator Angel Wong, PhD, from the London School of Hygiene and Tropical Medicine, said in an interview.

And with nearly 11 million NSAID prescriptions dispensed in primary care in England alone in the past 12 months, the inconsistency was concerning.

The team compared COVID-19–related deaths in two groups: one group of more than 700,000 people taking NSAIDs, including patients with rheumatoid arthritis and osteoarthritis; and another of almost 3.5 million people not on the medication.

NSAIDs work by inhibiting cyclooxygenase-1 and COX-2 enzymes in the body, which are crucial for the generation of prostaglandins. These lipid molecules play a role in inflammation and are blocked by NSAIDs.

The investigators found no evidence of a harmful effect of NSAIDs on COVID-19-related deaths; their results were published online Jan. 21 in the Annals of the Rheumatic Diseases.

The results, they pointed out, are in line with a Danish study that also showed no evidence of a higher risk for severe COVID-19 outcomes with NSAID use.

“It’s reassuring,” Dr. Wong said, “that patients can safely continue treatment.”
 

More new evidence

Dr. Wilen’s team found that SARS-CoV-2 infection stimulated COX-2 expression in human and mice cells. However, suppression of COX-2 by two commonly used NSAIDs, ibuprofen and meloxicam, had no effect on ACE2 expression, viral entry, or viral replication.

In their mouse model of SARS-CoV-2 infection, the investigators saw that NSAIDs impaired the production of proinflammatory cytokines and neutralizing antibodies. The findings suggest that NSAIDs influence COVID-19 outcomes by dampening the inflammatory response and production of protective antibodies, rather than modifying susceptibility to infection or viral replication.

Understanding the effect of NSAIDs on cytokine production is critical, Dr. Wilen pointed out, because they might be protective early in COVID-19 but pathologic at later stages.

Timing is crucial in the case of other immunomodulatory drugs. For example, dexamethasone lowers mortality in COVID-19 patients on respiratory support but is potentially harmful for those with milder disease.

There still is a lot to learn, Dr. Wilen acknowledged. “We may be seeing something similar going on with NSAIDs, where the timing of treatment is important.”

A version of this article first appeared on Medscape.com.

More people with fever and body aches are turning to NSAIDs to ease symptoms, but the drugs have come under new scrutiny as investigators work to determine whether they are a safe way to relieve the pain of COVID-19 vaccination or symptoms of the disease.

Early on in the pandemic, French health officials warned that NSAIDs, such as ibuprofen, could worsen coronavirus disease, and they recommended switching to acetaminophen instead.

The National Health Service in the United Kingdom followed with a similar recommendation for acetaminophen.

But the European Medicines Agency took a different approach, reporting “no scientific evidence” that NSAIDs could worsen COVID-19. The U.S. Food and Drug Administration also opted not to take a stance.

The debate prompted discussion on social media, with various reactions from around the world. It also inspired Craig Wilen, MD, PhD, from Yale University, New Haven, Conn., and associates to examine the effect of NSAIDs on COVID-19 infection and immune response. Their findings were published online Jan.20 in the Journal of Virology.

“It really bothered me that non–evidence-based decisions were driving the conversation,” Dr. Wilen said. “Millions of people are taking NSAIDs every day and clinical decisions about their care shouldn’t be made on a hypothesis.”

One theory is that NSAIDs alter susceptibility to infection by modifying ACE2. The drugs might also change the cell entry receptor for SARS-CoV-2, alter virus replication, or even modify the immune response.

British researchers, also questioning the safety of NSAIDs in patients with COVID-19, delved into National Health Service records to study two large groups of patients, some of whom were taking the pain relievers.

“We were watching the controversy and the lack of evidence and wanted to contribute,” lead investigator Angel Wong, PhD, from the London School of Hygiene and Tropical Medicine, said in an interview.

And with nearly 11 million NSAID prescriptions dispensed in primary care in England alone in the past 12 months, the inconsistency was concerning.

The team compared COVID-19–related deaths in two groups: one group of more than 700,000 people taking NSAIDs, including patients with rheumatoid arthritis and osteoarthritis; and another of almost 3.5 million people not on the medication.

NSAIDs work by inhibiting cyclooxygenase-1 and COX-2 enzymes in the body, which are crucial for the generation of prostaglandins. These lipid molecules play a role in inflammation and are blocked by NSAIDs.

The investigators found no evidence of a harmful effect of NSAIDs on COVID-19-related deaths; their results were published online Jan. 21 in the Annals of the Rheumatic Diseases.

The results, they pointed out, are in line with a Danish study that also showed no evidence of a higher risk for severe COVID-19 outcomes with NSAID use.

“It’s reassuring,” Dr. Wong said, “that patients can safely continue treatment.”
 

More new evidence

Dr. Wilen’s team found that SARS-CoV-2 infection stimulated COX-2 expression in human and mice cells. However, suppression of COX-2 by two commonly used NSAIDs, ibuprofen and meloxicam, had no effect on ACE2 expression, viral entry, or viral replication.

In their mouse model of SARS-CoV-2 infection, the investigators saw that NSAIDs impaired the production of proinflammatory cytokines and neutralizing antibodies. The findings suggest that NSAIDs influence COVID-19 outcomes by dampening the inflammatory response and production of protective antibodies, rather than modifying susceptibility to infection or viral replication.

Understanding the effect of NSAIDs on cytokine production is critical, Dr. Wilen pointed out, because they might be protective early in COVID-19 but pathologic at later stages.

Timing is crucial in the case of other immunomodulatory drugs. For example, dexamethasone lowers mortality in COVID-19 patients on respiratory support but is potentially harmful for those with milder disease.

There still is a lot to learn, Dr. Wilen acknowledged. “We may be seeing something similar going on with NSAIDs, where the timing of treatment is important.”

A version of this article first appeared on Medscape.com.

More people with fever and body aches are turning to NSAIDs to ease symptoms, but the drugs have come under new scrutiny as investigators work to determine whether they are a safe way to relieve the pain of COVID-19 vaccination or symptoms of the disease.

Early on in the pandemic, French health officials warned that NSAIDs, such as ibuprofen, could worsen coronavirus disease, and they recommended switching to acetaminophen instead.

The National Health Service in the United Kingdom followed with a similar recommendation for acetaminophen.

But the European Medicines Agency took a different approach, reporting “no scientific evidence” that NSAIDs could worsen COVID-19. The U.S. Food and Drug Administration also opted not to take a stance.

The debate prompted discussion on social media, with various reactions from around the world. It also inspired Craig Wilen, MD, PhD, from Yale University, New Haven, Conn., and associates to examine the effect of NSAIDs on COVID-19 infection and immune response. Their findings were published online Jan.20 in the Journal of Virology.

“It really bothered me that non–evidence-based decisions were driving the conversation,” Dr. Wilen said. “Millions of people are taking NSAIDs every day and clinical decisions about their care shouldn’t be made on a hypothesis.”

One theory is that NSAIDs alter susceptibility to infection by modifying ACE2. The drugs might also change the cell entry receptor for SARS-CoV-2, alter virus replication, or even modify the immune response.

British researchers, also questioning the safety of NSAIDs in patients with COVID-19, delved into National Health Service records to study two large groups of patients, some of whom were taking the pain relievers.

“We were watching the controversy and the lack of evidence and wanted to contribute,” lead investigator Angel Wong, PhD, from the London School of Hygiene and Tropical Medicine, said in an interview.

And with nearly 11 million NSAID prescriptions dispensed in primary care in England alone in the past 12 months, the inconsistency was concerning.

The team compared COVID-19–related deaths in two groups: one group of more than 700,000 people taking NSAIDs, including patients with rheumatoid arthritis and osteoarthritis; and another of almost 3.5 million people not on the medication.

NSAIDs work by inhibiting cyclooxygenase-1 and COX-2 enzymes in the body, which are crucial for the generation of prostaglandins. These lipid molecules play a role in inflammation and are blocked by NSAIDs.

The investigators found no evidence of a harmful effect of NSAIDs on COVID-19-related deaths; their results were published online Jan. 21 in the Annals of the Rheumatic Diseases.

The results, they pointed out, are in line with a Danish study that also showed no evidence of a higher risk for severe COVID-19 outcomes with NSAID use.

“It’s reassuring,” Dr. Wong said, “that patients can safely continue treatment.”
 

More new evidence

Dr. Wilen’s team found that SARS-CoV-2 infection stimulated COX-2 expression in human and mice cells. However, suppression of COX-2 by two commonly used NSAIDs, ibuprofen and meloxicam, had no effect on ACE2 expression, viral entry, or viral replication.

In their mouse model of SARS-CoV-2 infection, the investigators saw that NSAIDs impaired the production of proinflammatory cytokines and neutralizing antibodies. The findings suggest that NSAIDs influence COVID-19 outcomes by dampening the inflammatory response and production of protective antibodies, rather than modifying susceptibility to infection or viral replication.

Understanding the effect of NSAIDs on cytokine production is critical, Dr. Wilen pointed out, because they might be protective early in COVID-19 but pathologic at later stages.

Timing is crucial in the case of other immunomodulatory drugs. For example, dexamethasone lowers mortality in COVID-19 patients on respiratory support but is potentially harmful for those with milder disease.

There still is a lot to learn, Dr. Wilen acknowledged. “We may be seeing something similar going on with NSAIDs, where the timing of treatment is important.”

A version of this article first appeared on Medscape.com.

A 34-year-old woman with no significant past medical history presented as a new patient to our family medicine clinic with 2 weeks of intermittent lower abdominal and pelvic pain. She was sexually active with 1 partner and denied abnormal vaginal discharge or bleeding. She mentioned she’d had an intrauterine contraceptive device (IUD) placed a few weeks ago. The patient was afebrile, and her pelvic examination was unremarkable.

Physical examination showed mild tenderness to palpation over the lower abdomen without rebound tenderness or guarding. A complete metabolic panel revealed no significant abnormalities, and her human chorionic gonadotropin levels were normal.

Findings from the physical exam and her clinical history prompted the need for imaging. An abdominal radiograph (FIGURE 1) and noncontrast computed tomography (FIGURES 2A and 2B) were subsequently ordered.

WHAT IS YOUR DIAGNOSIS?
HOW WOULD YOU TREAT THIS PATIENT?

The abdominal radiograph revealed a nonobstructive bowel gas pattern with an IUD overlaying the central lower abdomen and pelvis at the L5-S1 level (FIGURE 1). Computed tomography (CT) of her abdomen and pelvis showed that the IUD was outside the endometrial cavity (FIGURES 2A and 2B). There was no evidence of pneumoperitoneum or bowel perforation. Based on the work-up and imaging, the patient’s pain was due to intra-abdominal IUD malpositioning.

Diagnostic criteria for IUD malpositioning include device migration into 1 of several locations, such as the lower uterine segment or cervix. IUD malpositioning can involve the rotation or protrusion of the device into or through the myometrium. On imaging, a well-positioned IUD should have a straight stem contained within the endometrial cavity, with the arms of the IUD extending laterally at the uterine fundus.

For our patient, an abdominal radiograph showed that her IUD was superiorly displaced outside the expected region of the endometrial cavity. CT helped to confirm this.

Complications with IUDs are few

Using an IUD is an increasingly popular method of contraception because it is effective and generally well tolerated, with minimal adverse effects or complications. In a multicenter retrospective chart review of 2138 patients who had IUDs, Aoun et al found that serious complications included pelvic inflammatory disease (2%), IUD expulsion (6%), and pregnancy (1%).¹ In a retrospective cohort study examining complications among 90,489 women with IUDs, Berenson et al found ectopic pregnancy and uterine perforation affected < 1%.²

A less serious complication is IUD malpositioning. Although it does seem to occur more often than other, more serious complications, the exact incidence is unknown. In a retrospective case-control study, Braaten et al reported the rate for IUD malpositioning was 10.4% among 182 women.³ Malpositioned IUDs may be more likely to occur in those with suspected adenomyosis.³ In a study by de Kroon et al, the estimated prevalence rate for an abnormal IUD position ranged from 4% to 7.7% among 195 patients.⁴

Continue to: The clinical presentation of IUD migration

Identification of a malpositioned IUD is needed to avoid the possible increased risk for uterine perforation, IUD expulsion, or pregnancy.⁵

IUDs that have perforated the uterus float freely in the pelvis or abdomen and can result in injury to adjacent structures as well as peritonitis, fistulas, and hemorrhage.^5-7 In addition, adhesion formation over the IUD can lead to intestinal obstruction, infertility, and chronic pain.⁶

Common symptoms of IUD malpositioning include abdominal or pelvic pain and abnormal bleeding, although many patients may be asymptomatic.⁸ In a retrospective study of 167 patients with IUDs who underwent pelvic ultrasound, 28 patients were found to have an IUD in an abnormal position.⁸ Rates of bleeding and pain were higher in patients with malpositioned IUDs (35.7% and 39.3%, respectively) than in those with a normally positioned IUD (15.1% and 19.4%, respectively).⁸

The differential Dx includes endometriosis and fibroids

IUD malpositioning can be distinguished from other diagnoses that cause pelvic pain and have similar presentations—including endometriosis, ectopic pregnancy, and fibroids—through imaging study findings, clinical history, and presentation.

Other conditions that may need to be ruled out include pelvic inflammatory disease, acute appendicitis, and ovarian cysts.⁹ A thorough history and physical examination can help rule out these conditions by organ system, and laboratory and imaging studies can help to confirm the diagnosis.

Continue to: Which imaging tool to use, and when

Assessment of intrauterine contraception placement requires evaluation of the uterine cavity; gynecologic examination alone is not sufficient to fully evaluate for IUD position. Certain imaging studies are particularly helpful for revealing possible IUD migration.

Ultrasound—a widely available, radiation-free modality—is the first-line imaging tool for evaluation of an IUD’s position.¹⁰ In addition, ultrasound can provide effective evaluation of other pelvic structures, which is helpful in identifying or eliminating other causes of pain or abnormal bleeding.

Conventional radiography. If the IUD is not visualized on ultrasound, the American College of Obstetricians and Gynecologists (ACOG) recommends radiography to determine if the IUD has been expelled or has migrated to an extra-uterine position.⁶

CT may be best suited for the evaluation of more severe complications of IUD malpositioning, including visceral perforation, abscess formation, or bowel obstruction. CT should be considered if the patient’s clinical presentation is suspicious for a more serious intra-abdominal pathology.

Management depends on the IUD’s position

For patients whose IUD has an uncertain position or nonvisualized intravaginal strings, ACOG’s first-line recommendations include ruling out pregnancy, using an alternative method for contraception, and ordering pelvic ultrasonography.⁶ ACOG recommendations for the management of IUD malpositioning depend on the device’s location and the patient’s symptomatology.

Continue to: Management of low-lying IUDs

Management of low-lying IUDs is complex. An IUD that is malpositioned in the cervix is considered partially expelled and should be completely removed.⁶ For asymptomatic patients with an IUD located in the lower uterine segment and above the internal cervical os, there should be strong consideration given to leaving the IUD in place because removal is associated with higher rates of pregnancy given the low rates of initiation of effective contraception following removal.⁶

For asymptomatic patients with an IUD in the lower uterine segment and above the internal cervical os, consider leaving the IUD in place.

IUD malpositioning in the peritoneal cavity requires surgical intervention. Although ACOG’s first-line recommendation is laparoscopic intervention, laparotomy can be considered if laparoscopy does not result in the removal of the IUD or the patient has more severe complications (sepsis or bowel perforation).⁶ At the time of IUD removal, the clinician should also discuss and/or prescribe interim contraception.

Treatment for our patient included uncomplicated laparoscopic surgical removal of the intra-abdominal IUD. The patient’s symptoms went away following the procedure, and she was subsequently switched to an oral contraceptive.

A 34-year-old woman with no significant past medical history presented as a new patient to our family medicine clinic with 2 weeks of intermittent lower abdominal and pelvic pain. She was sexually active with 1 partner and denied abnormal vaginal discharge or bleeding. She mentioned she’d had an intrauterine contraceptive device (IUD) placed a few weeks ago. The patient was afebrile, and her pelvic examination was unremarkable.

Physical examination showed mild tenderness to palpation over the lower abdomen without rebound tenderness or guarding. A complete metabolic panel revealed no significant abnormalities, and her human chorionic gonadotropin levels were normal.

Findings from the physical exam and her clinical history prompted the need for imaging. An abdominal radiograph (FIGURE 1) and noncontrast computed tomography (FIGURES 2A and 2B) were subsequently ordered.

WHAT IS YOUR DIAGNOSIS?
HOW WOULD YOU TREAT THIS PATIENT?

The abdominal radiograph revealed a nonobstructive bowel gas pattern with an IUD overlaying the central lower abdomen and pelvis at the L5-S1 level (FIGURE 1). Computed tomography (CT) of her abdomen and pelvis showed that the IUD was outside the endometrial cavity (FIGURES 2A and 2B). There was no evidence of pneumoperitoneum or bowel perforation. Based on the work-up and imaging, the patient’s pain was due to intra-abdominal IUD malpositioning.

Diagnostic criteria for IUD malpositioning include device migration into 1 of several locations, such as the lower uterine segment or cervix. IUD malpositioning can involve the rotation or protrusion of the device into or through the myometrium. On imaging, a well-positioned IUD should have a straight stem contained within the endometrial cavity, with the arms of the IUD extending laterally at the uterine fundus.

For our patient, an abdominal radiograph showed that her IUD was superiorly displaced outside the expected region of the endometrial cavity. CT helped to confirm this.

Complications with IUDs are few

Using an IUD is an increasingly popular method of contraception because it is effective and generally well tolerated, with minimal adverse effects or complications. In a multicenter retrospective chart review of 2138 patients who had IUDs, Aoun et al found that serious complications included pelvic inflammatory disease (2%), IUD expulsion (6%), and pregnancy (1%).¹ In a retrospective cohort study examining complications among 90,489 women with IUDs, Berenson et al found ectopic pregnancy and uterine perforation affected < 1%.²

A less serious complication is IUD malpositioning. Although it does seem to occur more often than other, more serious complications, the exact incidence is unknown. In a retrospective case-control study, Braaten et al reported the rate for IUD malpositioning was 10.4% among 182 women.³ Malpositioned IUDs may be more likely to occur in those with suspected adenomyosis.³ In a study by de Kroon et al, the estimated prevalence rate for an abnormal IUD position ranged from 4% to 7.7% among 195 patients.⁴

Continue to: The clinical presentation of IUD migration

Identification of a malpositioned IUD is needed to avoid the possible increased risk for uterine perforation, IUD expulsion, or pregnancy.⁵

IUDs that have perforated the uterus float freely in the pelvis or abdomen and can result in injury to adjacent structures as well as peritonitis, fistulas, and hemorrhage.^5-7 In addition, adhesion formation over the IUD can lead to intestinal obstruction, infertility, and chronic pain.⁶

Common symptoms of IUD malpositioning include abdominal or pelvic pain and abnormal bleeding, although many patients may be asymptomatic.⁸ In a retrospective study of 167 patients with IUDs who underwent pelvic ultrasound, 28 patients were found to have an IUD in an abnormal position.⁸ Rates of bleeding and pain were higher in patients with malpositioned IUDs (35.7% and 39.3%, respectively) than in those with a normally positioned IUD (15.1% and 19.4%, respectively).⁸

The differential Dx includes endometriosis and fibroids

IUD malpositioning can be distinguished from other diagnoses that cause pelvic pain and have similar presentations—including endometriosis, ectopic pregnancy, and fibroids—through imaging study findings, clinical history, and presentation.

Other conditions that may need to be ruled out include pelvic inflammatory disease, acute appendicitis, and ovarian cysts.⁹ A thorough history and physical examination can help rule out these conditions by organ system, and laboratory and imaging studies can help to confirm the diagnosis.

Continue to: Which imaging tool to use, and when

Assessment of intrauterine contraception placement requires evaluation of the uterine cavity; gynecologic examination alone is not sufficient to fully evaluate for IUD position. Certain imaging studies are particularly helpful for revealing possible IUD migration.

Ultrasound—a widely available, radiation-free modality—is the first-line imaging tool for evaluation of an IUD’s position.¹⁰ In addition, ultrasound can provide effective evaluation of other pelvic structures, which is helpful in identifying or eliminating other causes of pain or abnormal bleeding.

Conventional radiography. If the IUD is not visualized on ultrasound, the American College of Obstetricians and Gynecologists (ACOG) recommends radiography to determine if the IUD has been expelled or has migrated to an extra-uterine position.⁶

CT may be best suited for the evaluation of more severe complications of IUD malpositioning, including visceral perforation, abscess formation, or bowel obstruction. CT should be considered if the patient’s clinical presentation is suspicious for a more serious intra-abdominal pathology.

Management depends on the IUD’s position

For patients whose IUD has an uncertain position or nonvisualized intravaginal strings, ACOG’s first-line recommendations include ruling out pregnancy, using an alternative method for contraception, and ordering pelvic ultrasonography.⁶ ACOG recommendations for the management of IUD malpositioning depend on the device’s location and the patient’s symptomatology.

Continue to: Management of low-lying IUDs

Management of low-lying IUDs is complex. An IUD that is malpositioned in the cervix is considered partially expelled and should be completely removed.⁶ For asymptomatic patients with an IUD located in the lower uterine segment and above the internal cervical os, there should be strong consideration given to leaving the IUD in place because removal is associated with higher rates of pregnancy given the low rates of initiation of effective contraception following removal.⁶

For asymptomatic patients with an IUD in the lower uterine segment and above the internal cervical os, consider leaving the IUD in place.

IUD malpositioning in the peritoneal cavity requires surgical intervention. Although ACOG’s first-line recommendation is laparoscopic intervention, laparotomy can be considered if laparoscopy does not result in the removal of the IUD or the patient has more severe complications (sepsis or bowel perforation).⁶ At the time of IUD removal, the clinician should also discuss and/or prescribe interim contraception.

Treatment for our patient included uncomplicated laparoscopic surgical removal of the intra-abdominal IUD. The patient’s symptoms went away following the procedure, and she was subsequently switched to an oral contraceptive.

A 34-year-old woman with no significant past medical history presented as a new patient to our family medicine clinic with 2 weeks of intermittent lower abdominal and pelvic pain. She was sexually active with 1 partner and denied abnormal vaginal discharge or bleeding. She mentioned she’d had an intrauterine contraceptive device (IUD) placed a few weeks ago. The patient was afebrile, and her pelvic examination was unremarkable.

Physical examination showed mild tenderness to palpation over the lower abdomen without rebound tenderness or guarding. A complete metabolic panel revealed no significant abnormalities, and her human chorionic gonadotropin levels were normal.

Findings from the physical exam and her clinical history prompted the need for imaging. An abdominal radiograph (FIGURE 1) and noncontrast computed tomography (FIGURES 2A and 2B) were subsequently ordered.

WHAT IS YOUR DIAGNOSIS?
HOW WOULD YOU TREAT THIS PATIENT?

The abdominal radiograph revealed a nonobstructive bowel gas pattern with an IUD overlaying the central lower abdomen and pelvis at the L5-S1 level (FIGURE 1). Computed tomography (CT) of her abdomen and pelvis showed that the IUD was outside the endometrial cavity (FIGURES 2A and 2B). There was no evidence of pneumoperitoneum or bowel perforation. Based on the work-up and imaging, the patient’s pain was due to intra-abdominal IUD malpositioning.

Diagnostic criteria for IUD malpositioning include device migration into 1 of several locations, such as the lower uterine segment or cervix. IUD malpositioning can involve the rotation or protrusion of the device into or through the myometrium. On imaging, a well-positioned IUD should have a straight stem contained within the endometrial cavity, with the arms of the IUD extending laterally at the uterine fundus.

For our patient, an abdominal radiograph showed that her IUD was superiorly displaced outside the expected region of the endometrial cavity. CT helped to confirm this.

Complications with IUDs are few

Using an IUD is an increasingly popular method of contraception because it is effective and generally well tolerated, with minimal adverse effects or complications. In a multicenter retrospective chart review of 2138 patients who had IUDs, Aoun et al found that serious complications included pelvic inflammatory disease (2%), IUD expulsion (6%), and pregnancy (1%).¹ In a retrospective cohort study examining complications among 90,489 women with IUDs, Berenson et al found ectopic pregnancy and uterine perforation affected < 1%.²

A less serious complication is IUD malpositioning. Although it does seem to occur more often than other, more serious complications, the exact incidence is unknown. In a retrospective case-control study, Braaten et al reported the rate for IUD malpositioning was 10.4% among 182 women.³ Malpositioned IUDs may be more likely to occur in those with suspected adenomyosis.³ In a study by de Kroon et al, the estimated prevalence rate for an abnormal IUD position ranged from 4% to 7.7% among 195 patients.⁴

Continue to: The clinical presentation of IUD migration

Identification of a malpositioned IUD is needed to avoid the possible increased risk for uterine perforation, IUD expulsion, or pregnancy.⁵

IUDs that have perforated the uterus float freely in the pelvis or abdomen and can result in injury to adjacent structures as well as peritonitis, fistulas, and hemorrhage.^5-7 In addition, adhesion formation over the IUD can lead to intestinal obstruction, infertility, and chronic pain.⁶

Common symptoms of IUD malpositioning include abdominal or pelvic pain and abnormal bleeding, although many patients may be asymptomatic.⁸ In a retrospective study of 167 patients with IUDs who underwent pelvic ultrasound, 28 patients were found to have an IUD in an abnormal position.⁸ Rates of bleeding and pain were higher in patients with malpositioned IUDs (35.7% and 39.3%, respectively) than in those with a normally positioned IUD (15.1% and 19.4%, respectively).⁸

The differential Dx includes endometriosis and fibroids

IUD malpositioning can be distinguished from other diagnoses that cause pelvic pain and have similar presentations—including endometriosis, ectopic pregnancy, and fibroids—through imaging study findings, clinical history, and presentation.

Other conditions that may need to be ruled out include pelvic inflammatory disease, acute appendicitis, and ovarian cysts.⁹ A thorough history and physical examination can help rule out these conditions by organ system, and laboratory and imaging studies can help to confirm the diagnosis.

Continue to: Which imaging tool to use, and when

Assessment of intrauterine contraception placement requires evaluation of the uterine cavity; gynecologic examination alone is not sufficient to fully evaluate for IUD position. Certain imaging studies are particularly helpful for revealing possible IUD migration.

Ultrasound—a widely available, radiation-free modality—is the first-line imaging tool for evaluation of an IUD’s position.¹⁰ In addition, ultrasound can provide effective evaluation of other pelvic structures, which is helpful in identifying or eliminating other causes of pain or abnormal bleeding.

Conventional radiography. If the IUD is not visualized on ultrasound, the American College of Obstetricians and Gynecologists (ACOG) recommends radiography to determine if the IUD has been expelled or has migrated to an extra-uterine position.⁶

CT may be best suited for the evaluation of more severe complications of IUD malpositioning, including visceral perforation, abscess formation, or bowel obstruction. CT should be considered if the patient’s clinical presentation is suspicious for a more serious intra-abdominal pathology.

Management depends on the IUD’s position

For patients whose IUD has an uncertain position or nonvisualized intravaginal strings, ACOG’s first-line recommendations include ruling out pregnancy, using an alternative method for contraception, and ordering pelvic ultrasonography.⁶ ACOG recommendations for the management of IUD malpositioning depend on the device’s location and the patient’s symptomatology.

Continue to: Management of low-lying IUDs

Management of low-lying IUDs is complex. An IUD that is malpositioned in the cervix is considered partially expelled and should be completely removed.⁶ For asymptomatic patients with an IUD located in the lower uterine segment and above the internal cervical os, there should be strong consideration given to leaving the IUD in place because removal is associated with higher rates of pregnancy given the low rates of initiation of effective contraception following removal.⁶

For asymptomatic patients with an IUD in the lower uterine segment and above the internal cervical os, consider leaving the IUD in place.

IUD malpositioning in the peritoneal cavity requires surgical intervention. Although ACOG’s first-line recommendation is laparoscopic intervention, laparotomy can be considered if laparoscopy does not result in the removal of the IUD or the patient has more severe complications (sepsis or bowel perforation).⁶ At the time of IUD removal, the clinician should also discuss and/or prescribe interim contraception.

Treatment for our patient included uncomplicated laparoscopic surgical removal of the intra-abdominal IUD. The patient’s symptoms went away following the procedure, and she was subsequently switched to an oral contraceptive.

THE CASE

A 60-year-old man with a past medical history of gastroesophageal reflux disease (GERD) and dyslipidemia presented to his family physician for evaluation of chronic cough. Five years prior, the patient had developed a high fever and respiratory symptoms, including a cough, and was believed to have had severe otitis media. He was treated with multiple courses of antibiotics and corticosteroids for persistent otitis media. Although the condition eventually resolved, his cough continued.

The persistent cough prompted the patient to consult a succession of specialists. First, he saw a gastroenterologist; following an esophagogastroduodenoscopy, he was prescribed pantoprazole. Despite the proton-pump inhibitor (PPI) therapy, the cough remained. Next, he had multiple visits with an otolaryngologist but that yielded no specific diagnosis for the cough. He also saw an allergist-immunologist, who identified a ragweed allergy, gave him a diagnosis of cough-variant asthma, and prescribed antihistamines and mometasone furoate and formoterol fumarate dihydrate. Neither was helpful.

After 5 years of frustration, the patient complained to his family physician that he still had a cough and “a tickle” in his throat that was worsened by speaking and drinking cold beverages. He denied fever, shortness of breath, nausea, vomiting, or any other associated symptoms.

THE DIAGNOSIS

The failed treatment attempts with antihistamines, corticosteroids, bronchodilators, and PPI therapy excluded multiple etiologies for the cough. The throat discomfort and feeling of a “tickle” prompted us to consider a nerve-related disorder on the differential. The diagnosis of laryngeal sensory neuropathy (LSN) was considered.

DISCUSSION

LSN is a relatively uncommon cause of chronic refractory cough that can also manifest with throat discomfort, dysphagia, and dysphonia.¹ It is thought to result from some type of insult to the recurrent laryngeal nerve or superior laryngeal nerve via viral infections, metabolic changes, or mechanical trauma, leading to a change in the firing threshold.² The hypothesis of nerve damage is supported by the increased incidence of LSN in patients with goiters and those with type 2 diabetes.^3,4 When there is a decrease in the laryngeal sensory threshold, dysfunctional laryngeal behavior results, leading to symptoms such as persistent cough and throat clearing.

Diagnosis. LSN is often diagnosed clinically, after GERD, allergies, asthma, angiotensin-converting enzyme inhibitor intake, and psychogenic disorders have been ruled out.¹ Our patient had a prior diagnosis or investigation of nearly all of these conditions. Other clues pointing to an LSN diagnosis include a cough lasting 8 weeks or more, recurrent sensory disturbances (such as a tickle) of instantaneous onset before each cough episode, triggers that can include talking or a change in air temperature, daily coughing episodes numbering in the 10s to 100s, and a nonproductive cough.^5,6

The throat discomfort and feeling of a “tickle” prompted us to consider a nerve-related disorder on the differential.

Beyond clinical clues, laryngeal electromyography, which evaluates the neuromuscular system in the larynx by recording action potentials generated in the laryngeal muscles during contraction, can be used for diagnosis.⁴ Videostroboscopy, which allows for an enlarged and slow motion view of the vocal cords, can also be used.

Continue to: Treatment

Treatment. To both confirm the diagnosis and treat the patient in a rapid, practical fashion, a trial of a neuromodulating agent such as pregabalin or gabapentin can be employed.^6-9 A study identifying 28 LSN patients found symptomatic relief in 68% of patients taking gabapentin 100 to 900 mg/d.² In another study, 12 LSN patients given pregabalin found relief after a 1-month regimen.¹ Another study of 12 patients showed amitriptyline hydrochloride and gabapentin provided a positive response in 2 months, and the addition of reflux precautions and acid-­suppression therapy was helpful.⁹ Finally, a group of 32 patients trialed on 3 different medications (amitriptyline, desipramine, and gabapentin) found similar efficacy among the 3.⁶

Another option. Aside from medications, botulinum toxin type A has been shown in a case series to directly decrease laryngeal hypertonicity and possibly reduce neurogenic inflammation and neuropeptide-mediated cough.¹⁰ Another study found that 18 patients with neurogenic cough who received superior laryngeal nerve blocks had cough severity index scores decrease from an average of 26.8 pretreatment to 14.6 posttreatment (P < .0001).¹¹

Our patient agreed to a trial of gabapentin 300 mg once a day, with titration up to a maximum of 900 mg tid. When the patient returned to the clinic 4 months later, he reported that when he reached 300 mg bid, the cough completely resolved.

THE TAKEAWAY

A persistent cough with minimal identifiable triggers is a huge disruption to a patient’s life; having to visit multiple specialists before receiving a diagnosis compounds that. In our patient’s case, the process took 5 years, which underscores how important it is that LSN be considered in the differential diagnosis. Since this is generally a diagnosis of exclusion, it is important to take a careful history of a patient with a chronic cough. If LSN seems likely, trialing a patient on neuromodulating medication is the next best step, with dose titration if necessary.

CORRESPONDENCE
Selena R. Pasadyn, 675 West 130th Street, Hinckley, OH, 44233; [email protected]

THE CASE

A 60-year-old man with a past medical history of gastroesophageal reflux disease (GERD) and dyslipidemia presented to his family physician for evaluation of chronic cough. Five years prior, the patient had developed a high fever and respiratory symptoms, including a cough, and was believed to have had severe otitis media. He was treated with multiple courses of antibiotics and corticosteroids for persistent otitis media. Although the condition eventually resolved, his cough continued.

The persistent cough prompted the patient to consult a succession of specialists. First, he saw a gastroenterologist; following an esophagogastroduodenoscopy, he was prescribed pantoprazole. Despite the proton-pump inhibitor (PPI) therapy, the cough remained. Next, he had multiple visits with an otolaryngologist but that yielded no specific diagnosis for the cough. He also saw an allergist-immunologist, who identified a ragweed allergy, gave him a diagnosis of cough-variant asthma, and prescribed antihistamines and mometasone furoate and formoterol fumarate dihydrate. Neither was helpful.

After 5 years of frustration, the patient complained to his family physician that he still had a cough and “a tickle” in his throat that was worsened by speaking and drinking cold beverages. He denied fever, shortness of breath, nausea, vomiting, or any other associated symptoms.

THE DIAGNOSIS

The failed treatment attempts with antihistamines, corticosteroids, bronchodilators, and PPI therapy excluded multiple etiologies for the cough. The throat discomfort and feeling of a “tickle” prompted us to consider a nerve-related disorder on the differential. The diagnosis of laryngeal sensory neuropathy (LSN) was considered.

DISCUSSION

LSN is a relatively uncommon cause of chronic refractory cough that can also manifest with throat discomfort, dysphagia, and dysphonia.¹ It is thought to result from some type of insult to the recurrent laryngeal nerve or superior laryngeal nerve via viral infections, metabolic changes, or mechanical trauma, leading to a change in the firing threshold.² The hypothesis of nerve damage is supported by the increased incidence of LSN in patients with goiters and those with type 2 diabetes.^3,4 When there is a decrease in the laryngeal sensory threshold, dysfunctional laryngeal behavior results, leading to symptoms such as persistent cough and throat clearing.

Diagnosis. LSN is often diagnosed clinically, after GERD, allergies, asthma, angiotensin-converting enzyme inhibitor intake, and psychogenic disorders have been ruled out.¹ Our patient had a prior diagnosis or investigation of nearly all of these conditions. Other clues pointing to an LSN diagnosis include a cough lasting 8 weeks or more, recurrent sensory disturbances (such as a tickle) of instantaneous onset before each cough episode, triggers that can include talking or a change in air temperature, daily coughing episodes numbering in the 10s to 100s, and a nonproductive cough.^5,6

The throat discomfort and feeling of a “tickle” prompted us to consider a nerve-related disorder on the differential.

Beyond clinical clues, laryngeal electromyography, which evaluates the neuromuscular system in the larynx by recording action potentials generated in the laryngeal muscles during contraction, can be used for diagnosis.⁴ Videostroboscopy, which allows for an enlarged and slow motion view of the vocal cords, can also be used.

Continue to: Treatment

Treatment. To both confirm the diagnosis and treat the patient in a rapid, practical fashion, a trial of a neuromodulating agent such as pregabalin or gabapentin can be employed.^6-9 A study identifying 28 LSN patients found symptomatic relief in 68% of patients taking gabapentin 100 to 900 mg/d.² In another study, 12 LSN patients given pregabalin found relief after a 1-month regimen.¹ Another study of 12 patients showed amitriptyline hydrochloride and gabapentin provided a positive response in 2 months, and the addition of reflux precautions and acid-­suppression therapy was helpful.⁹ Finally, a group of 32 patients trialed on 3 different medications (amitriptyline, desipramine, and gabapentin) found similar efficacy among the 3.⁶

Another option. Aside from medications, botulinum toxin type A has been shown in a case series to directly decrease laryngeal hypertonicity and possibly reduce neurogenic inflammation and neuropeptide-mediated cough.¹⁰ Another study found that 18 patients with neurogenic cough who received superior laryngeal nerve blocks had cough severity index scores decrease from an average of 26.8 pretreatment to 14.6 posttreatment (P < .0001).¹¹

Our patient agreed to a trial of gabapentin 300 mg once a day, with titration up to a maximum of 900 mg tid. When the patient returned to the clinic 4 months later, he reported that when he reached 300 mg bid, the cough completely resolved.

THE TAKEAWAY

A persistent cough with minimal identifiable triggers is a huge disruption to a patient’s life; having to visit multiple specialists before receiving a diagnosis compounds that. In our patient’s case, the process took 5 years, which underscores how important it is that LSN be considered in the differential diagnosis. Since this is generally a diagnosis of exclusion, it is important to take a careful history of a patient with a chronic cough. If LSN seems likely, trialing a patient on neuromodulating medication is the next best step, with dose titration if necessary.

CORRESPONDENCE
Selena R. Pasadyn, 675 West 130th Street, Hinckley, OH, 44233; [email protected]

THE CASE

A 60-year-old man with a past medical history of gastroesophageal reflux disease (GERD) and dyslipidemia presented to his family physician for evaluation of chronic cough. Five years prior, the patient had developed a high fever and respiratory symptoms, including a cough, and was believed to have had severe otitis media. He was treated with multiple courses of antibiotics and corticosteroids for persistent otitis media. Although the condition eventually resolved, his cough continued.

The persistent cough prompted the patient to consult a succession of specialists. First, he saw a gastroenterologist; following an esophagogastroduodenoscopy, he was prescribed pantoprazole. Despite the proton-pump inhibitor (PPI) therapy, the cough remained. Next, he had multiple visits with an otolaryngologist but that yielded no specific diagnosis for the cough. He also saw an allergist-immunologist, who identified a ragweed allergy, gave him a diagnosis of cough-variant asthma, and prescribed antihistamines and mometasone furoate and formoterol fumarate dihydrate. Neither was helpful.

After 5 years of frustration, the patient complained to his family physician that he still had a cough and “a tickle” in his throat that was worsened by speaking and drinking cold beverages. He denied fever, shortness of breath, nausea, vomiting, or any other associated symptoms.

THE DIAGNOSIS

The failed treatment attempts with antihistamines, corticosteroids, bronchodilators, and PPI therapy excluded multiple etiologies for the cough. The throat discomfort and feeling of a “tickle” prompted us to consider a nerve-related disorder on the differential. The diagnosis of laryngeal sensory neuropathy (LSN) was considered.

DISCUSSION

LSN is a relatively uncommon cause of chronic refractory cough that can also manifest with throat discomfort, dysphagia, and dysphonia.¹ It is thought to result from some type of insult to the recurrent laryngeal nerve or superior laryngeal nerve via viral infections, metabolic changes, or mechanical trauma, leading to a change in the firing threshold.² The hypothesis of nerve damage is supported by the increased incidence of LSN in patients with goiters and those with type 2 diabetes.^3,4 When there is a decrease in the laryngeal sensory threshold, dysfunctional laryngeal behavior results, leading to symptoms such as persistent cough and throat clearing.

Diagnosis. LSN is often diagnosed clinically, after GERD, allergies, asthma, angiotensin-converting enzyme inhibitor intake, and psychogenic disorders have been ruled out.¹ Our patient had a prior diagnosis or investigation of nearly all of these conditions. Other clues pointing to an LSN diagnosis include a cough lasting 8 weeks or more, recurrent sensory disturbances (such as a tickle) of instantaneous onset before each cough episode, triggers that can include talking or a change in air temperature, daily coughing episodes numbering in the 10s to 100s, and a nonproductive cough.^5,6

The throat discomfort and feeling of a “tickle” prompted us to consider a nerve-related disorder on the differential.

Beyond clinical clues, laryngeal electromyography, which evaluates the neuromuscular system in the larynx by recording action potentials generated in the laryngeal muscles during contraction, can be used for diagnosis.⁴ Videostroboscopy, which allows for an enlarged and slow motion view of the vocal cords, can also be used.

Continue to: Treatment

Treatment. To both confirm the diagnosis and treat the patient in a rapid, practical fashion, a trial of a neuromodulating agent such as pregabalin or gabapentin can be employed.^6-9 A study identifying 28 LSN patients found symptomatic relief in 68% of patients taking gabapentin 100 to 900 mg/d.² In another study, 12 LSN patients given pregabalin found relief after a 1-month regimen.¹ Another study of 12 patients showed amitriptyline hydrochloride and gabapentin provided a positive response in 2 months, and the addition of reflux precautions and acid-­suppression therapy was helpful.⁹ Finally, a group of 32 patients trialed on 3 different medications (amitriptyline, desipramine, and gabapentin) found similar efficacy among the 3.⁶

Another option. Aside from medications, botulinum toxin type A has been shown in a case series to directly decrease laryngeal hypertonicity and possibly reduce neurogenic inflammation and neuropeptide-mediated cough.¹⁰ Another study found that 18 patients with neurogenic cough who received superior laryngeal nerve blocks had cough severity index scores decrease from an average of 26.8 pretreatment to 14.6 posttreatment (P < .0001).¹¹

Our patient agreed to a trial of gabapentin 300 mg once a day, with titration up to a maximum of 900 mg tid. When the patient returned to the clinic 4 months later, he reported that when he reached 300 mg bid, the cough completely resolved.

THE TAKEAWAY

A persistent cough with minimal identifiable triggers is a huge disruption to a patient’s life; having to visit multiple specialists before receiving a diagnosis compounds that. In our patient’s case, the process took 5 years, which underscores how important it is that LSN be considered in the differential diagnosis. Since this is generally a diagnosis of exclusion, it is important to take a careful history of a patient with a chronic cough. If LSN seems likely, trialing a patient on neuromodulating medication is the next best step, with dose titration if necessary.

CORRESPONDENCE
Selena R. Pasadyn, 675 West 130th Street, Hinckley, OH, 44233; [email protected]

THE CASE

A 20-year-old man presented to our clinic with a 3-day history of nonradiating chest pain located at the center of his chest. Past medical history included idiopathic neonatal giant-cell hepatitis and subsequent liver transplant at 1 month of age; he had been followed by the transplant team without rejection or infection and was in otherwise good health prior to the chest pain.

On the day of symptom onset, he was walking inside his house and fell to his knees with a chest pain described as “a punch” to the center of the chest that lasted for a few seconds. He was able to continue his daily activities without limitation despite a constant, squeezing, centrally located chest pain. The pain worsened with cough and exertion.

A few hours later, he went to an urgent care center for evaluation. There, he reported, his chest radiograph and electrocardiogram (EKG) results were normal and he was given a diagnosis of musculoskeletal chest pain. Over the next 3 days, his chest pain persisted but did not worsen. He was taking 500 mg of naproxen every 8 hours with no improvement. No other acute or chronic medications were being taken. He had no significant family history. A review of systems was otherwise negative.

On physical exam, his vital statistics included a height of 6’4”; weight, 261 lb; body mass index, 31.8; temperature, 98.7 °F; blood pressure, 134/77 mm Hg; heart rate, 92 beats/min; respiratory rate, 18 breaths/min; and oxygen saturation, 96%. Throughout the exam, he demonstrated no acute distress, appeared well, and was talkative; however, he reported having a “constant, squeezing” chest pain that did not worsen with palpation of the chest. The rest of his physical exam was unremarkable.

Although he reported that his EKG and chest radiograph were normal 3 days prior, repeat chest radiograph and EKG were ordered due to his unexplained, active chest pain and the lack of immediate access to the prior results.

THE DIAGNOSIS

The chest radiograph (FIGURE 1A) showed a “mildly ectatic ascending thoracic aorta” that had increased since a chest radiograph from 6 years prior (FIGURE 1B) and “was concerning for an aneurysm.” Computed tomography (CT) angiography (FIGURE 2) then confirmed a 7-cm aneurysm of the ascending aorta, with findings suggestive of a retrograde ascending aortic dissection.

DISCUSSION

The average age of a patient with acute aortic dissection (AAD) is 63 years; only 7% occur in people younger than 40.¹ AAD is often accompanied by a predisposing risk factor such as a connective tissue disease, bicuspid aortic valve, longstanding hypertension, trauma, or larger aortic dimensions.^2,3 Younger patients are more likely to have predisposing risk factors of Marfan syndrome, prior aortic surgery, or a bicuspid aortic valve.³

Continue to: A literature review did not reveal...

A literature review did not reveal any known correlation between the patient’s history of giant-cell hepatitis or antirejection therapy with thoracic aortic dissection. Furthermore, liver transplant is not known to be a specific risk factor for AAD in pediatric patients or outside the immediate postoperative period. Therefore, there were no known predisposing risk factors for AAD in our patient.

The most common clinical feature of AAD is chest pain, which occurs in 75% of patients.¹ Other clinical symptoms include hypertension and diaphoresis.^2,4 However, classic clinical findings are not always displayed, making the diagnosis difficult.^2,4 The classical description of “tearing pain” is seen in only 51% of patients, and 5% to 15% of patients present without any pain.¹

Commonly missed or misdiagnosed. The diagnosis of AAD has been missed during the initial exam in 38% of patients.⁴ As seen in our case, symptoms may be initially diagnosed as musculoskeletal chest pain. Based on symptoms, AAD can be incorrectly diagnosed as an acute myocardial infarction or vascular embolization.^2,4

Every hour after symptom onset, the mortality rate of untreated AAD increases 1% to 2%,with no difference based on age.^3,4 Different reports have shown mortality rates between 7% and 30%.⁴

Effective imaging is crucial to the diagnosis and treatment of AAD, given the occurrence of atypical presentation, missed diagnosis, and high mortality rate.⁴ A chest radiograph will show a widened mediastinum, but the preferred diagnostic tests are a CT or transthoracic echocardiogram.^2,4 Once the diagnosis of AAD is confirmed, an aortic angiogram is the preferred test to determine the extent of the dissection prior to surgical treatment.²

Continue to: Classification dictates treatment

Classification dictates treatment. AAD is classified based on where the dissection of the aorta occurs. If the dissection involves the ascending aorta, it is classified as a type A AAD and should immediately be treated with emergent surgery in order to prevent complications including myocardial infarction, cardiac tamponade, and aortic rupture.^2,4,5 If the dissection is limited to the descending aorta, it is classified as a type B AAD and can be medically managed by controlling pain and lowering blood pressure; if symptoms persist, surgical management may be required.² After hospital discharge, AAD patients are followed closely with medical therapy, serial imaging, and reoperation if necessary.⁴

Our patient underwent emergent surgery for aortic root/ascending aortic replacement with a mechanical valve. He tolerated the procedure well. Surgical tissue pathology of the aortic segment showed a wall of elastic vessel with medial degeneration and dissection, and the tissue pathology of the aorta leaflets showed valvular tissue with myxoid degeneration.

THE TAKEAWAY

It is critical to keep AAD in the differential diagnosis of a patient presenting with acute onset of chest pain, as AAD often has an atypical presentation and can easily be misdiagnosed. Effective imaging is crucial to diagnosis, and immediate treatment is essential to patient survival.

CORRESPONDENCE
Rachel A. Reedy, PA, University of Florida, Department of General Pediatrics, 7046 SW Archer Road, Gainesville, FL 32608; [email protected]

THE CASE

A 20-year-old man presented to our clinic with a 3-day history of nonradiating chest pain located at the center of his chest. Past medical history included idiopathic neonatal giant-cell hepatitis and subsequent liver transplant at 1 month of age; he had been followed by the transplant team without rejection or infection and was in otherwise good health prior to the chest pain.

On the day of symptom onset, he was walking inside his house and fell to his knees with a chest pain described as “a punch” to the center of the chest that lasted for a few seconds. He was able to continue his daily activities without limitation despite a constant, squeezing, centrally located chest pain. The pain worsened with cough and exertion.

A few hours later, he went to an urgent care center for evaluation. There, he reported, his chest radiograph and electrocardiogram (EKG) results were normal and he was given a diagnosis of musculoskeletal chest pain. Over the next 3 days, his chest pain persisted but did not worsen. He was taking 500 mg of naproxen every 8 hours with no improvement. No other acute or chronic medications were being taken. He had no significant family history. A review of systems was otherwise negative.

On physical exam, his vital statistics included a height of 6’4”; weight, 261 lb; body mass index, 31.8; temperature, 98.7 °F; blood pressure, 134/77 mm Hg; heart rate, 92 beats/min; respiratory rate, 18 breaths/min; and oxygen saturation, 96%. Throughout the exam, he demonstrated no acute distress, appeared well, and was talkative; however, he reported having a “constant, squeezing” chest pain that did not worsen with palpation of the chest. The rest of his physical exam was unremarkable.

Although he reported that his EKG and chest radiograph were normal 3 days prior, repeat chest radiograph and EKG were ordered due to his unexplained, active chest pain and the lack of immediate access to the prior results.

THE DIAGNOSIS

The chest radiograph (FIGURE 1A) showed a “mildly ectatic ascending thoracic aorta” that had increased since a chest radiograph from 6 years prior (FIGURE 1B) and “was concerning for an aneurysm.” Computed tomography (CT) angiography (FIGURE 2) then confirmed a 7-cm aneurysm of the ascending aorta, with findings suggestive of a retrograde ascending aortic dissection.

DISCUSSION

The average age of a patient with acute aortic dissection (AAD) is 63 years; only 7% occur in people younger than 40.¹ AAD is often accompanied by a predisposing risk factor such as a connective tissue disease, bicuspid aortic valve, longstanding hypertension, trauma, or larger aortic dimensions.^2,3 Younger patients are more likely to have predisposing risk factors of Marfan syndrome, prior aortic surgery, or a bicuspid aortic valve.³

Continue to: A literature review did not reveal...

A literature review did not reveal any known correlation between the patient’s history of giant-cell hepatitis or antirejection therapy with thoracic aortic dissection. Furthermore, liver transplant is not known to be a specific risk factor for AAD in pediatric patients or outside the immediate postoperative period. Therefore, there were no known predisposing risk factors for AAD in our patient.

The most common clinical feature of AAD is chest pain, which occurs in 75% of patients.¹ Other clinical symptoms include hypertension and diaphoresis.^2,4 However, classic clinical findings are not always displayed, making the diagnosis difficult.^2,4 The classical description of “tearing pain” is seen in only 51% of patients, and 5% to 15% of patients present without any pain.¹

Commonly missed or misdiagnosed. The diagnosis of AAD has been missed during the initial exam in 38% of patients.⁴ As seen in our case, symptoms may be initially diagnosed as musculoskeletal chest pain. Based on symptoms, AAD can be incorrectly diagnosed as an acute myocardial infarction or vascular embolization.^2,4

Every hour after symptom onset, the mortality rate of untreated AAD increases 1% to 2%,with no difference based on age.^3,4 Different reports have shown mortality rates between 7% and 30%.⁴

Effective imaging is crucial to the diagnosis and treatment of AAD, given the occurrence of atypical presentation, missed diagnosis, and high mortality rate.⁴ A chest radiograph will show a widened mediastinum, but the preferred diagnostic tests are a CT or transthoracic echocardiogram.^2,4 Once the diagnosis of AAD is confirmed, an aortic angiogram is the preferred test to determine the extent of the dissection prior to surgical treatment.²

Continue to: Classification dictates treatment

Classification dictates treatment. AAD is classified based on where the dissection of the aorta occurs. If the dissection involves the ascending aorta, it is classified as a type A AAD and should immediately be treated with emergent surgery in order to prevent complications including myocardial infarction, cardiac tamponade, and aortic rupture.^2,4,5 If the dissection is limited to the descending aorta, it is classified as a type B AAD and can be medically managed by controlling pain and lowering blood pressure; if symptoms persist, surgical management may be required.² After hospital discharge, AAD patients are followed closely with medical therapy, serial imaging, and reoperation if necessary.⁴

Our patient underwent emergent surgery for aortic root/ascending aortic replacement with a mechanical valve. He tolerated the procedure well. Surgical tissue pathology of the aortic segment showed a wall of elastic vessel with medial degeneration and dissection, and the tissue pathology of the aorta leaflets showed valvular tissue with myxoid degeneration.

THE TAKEAWAY

It is critical to keep AAD in the differential diagnosis of a patient presenting with acute onset of chest pain, as AAD often has an atypical presentation and can easily be misdiagnosed. Effective imaging is crucial to diagnosis, and immediate treatment is essential to patient survival.

CORRESPONDENCE
Rachel A. Reedy, PA, University of Florida, Department of General Pediatrics, 7046 SW Archer Road, Gainesville, FL 32608; [email protected]

THE CASE

A 20-year-old man presented to our clinic with a 3-day history of nonradiating chest pain located at the center of his chest. Past medical history included idiopathic neonatal giant-cell hepatitis and subsequent liver transplant at 1 month of age; he had been followed by the transplant team without rejection or infection and was in otherwise good health prior to the chest pain.

On the day of symptom onset, he was walking inside his house and fell to his knees with a chest pain described as “a punch” to the center of the chest that lasted for a few seconds. He was able to continue his daily activities without limitation despite a constant, squeezing, centrally located chest pain. The pain worsened with cough and exertion.

A few hours later, he went to an urgent care center for evaluation. There, he reported, his chest radiograph and electrocardiogram (EKG) results were normal and he was given a diagnosis of musculoskeletal chest pain. Over the next 3 days, his chest pain persisted but did not worsen. He was taking 500 mg of naproxen every 8 hours with no improvement. No other acute or chronic medications were being taken. He had no significant family history. A review of systems was otherwise negative.

On physical exam, his vital statistics included a height of 6’4”; weight, 261 lb; body mass index, 31.8; temperature, 98.7 °F; blood pressure, 134/77 mm Hg; heart rate, 92 beats/min; respiratory rate, 18 breaths/min; and oxygen saturation, 96%. Throughout the exam, he demonstrated no acute distress, appeared well, and was talkative; however, he reported having a “constant, squeezing” chest pain that did not worsen with palpation of the chest. The rest of his physical exam was unremarkable.

Although he reported that his EKG and chest radiograph were normal 3 days prior, repeat chest radiograph and EKG were ordered due to his unexplained, active chest pain and the lack of immediate access to the prior results.

THE DIAGNOSIS

The chest radiograph (FIGURE 1A) showed a “mildly ectatic ascending thoracic aorta” that had increased since a chest radiograph from 6 years prior (FIGURE 1B) and “was concerning for an aneurysm.” Computed tomography (CT) angiography (FIGURE 2) then confirmed a 7-cm aneurysm of the ascending aorta, with findings suggestive of a retrograde ascending aortic dissection.

DISCUSSION

The average age of a patient with acute aortic dissection (AAD) is 63 years; only 7% occur in people younger than 40.¹ AAD is often accompanied by a predisposing risk factor such as a connective tissue disease, bicuspid aortic valve, longstanding hypertension, trauma, or larger aortic dimensions.^2,3 Younger patients are more likely to have predisposing risk factors of Marfan syndrome, prior aortic surgery, or a bicuspid aortic valve.³

Continue to: A literature review did not reveal...

A literature review did not reveal any known correlation between the patient’s history of giant-cell hepatitis or antirejection therapy with thoracic aortic dissection. Furthermore, liver transplant is not known to be a specific risk factor for AAD in pediatric patients or outside the immediate postoperative period. Therefore, there were no known predisposing risk factors for AAD in our patient.

The most common clinical feature of AAD is chest pain, which occurs in 75% of patients.¹ Other clinical symptoms include hypertension and diaphoresis.^2,4 However, classic clinical findings are not always displayed, making the diagnosis difficult.^2,4 The classical description of “tearing pain” is seen in only 51% of patients, and 5% to 15% of patients present without any pain.¹

Commonly missed or misdiagnosed. The diagnosis of AAD has been missed during the initial exam in 38% of patients.⁴ As seen in our case, symptoms may be initially diagnosed as musculoskeletal chest pain. Based on symptoms, AAD can be incorrectly diagnosed as an acute myocardial infarction or vascular embolization.^2,4

Every hour after symptom onset, the mortality rate of untreated AAD increases 1% to 2%,with no difference based on age.^3,4 Different reports have shown mortality rates between 7% and 30%.⁴

Effective imaging is crucial to the diagnosis and treatment of AAD, given the occurrence of atypical presentation, missed diagnosis, and high mortality rate.⁴ A chest radiograph will show a widened mediastinum, but the preferred diagnostic tests are a CT or transthoracic echocardiogram.^2,4 Once the diagnosis of AAD is confirmed, an aortic angiogram is the preferred test to determine the extent of the dissection prior to surgical treatment.²

Continue to: Classification dictates treatment

Classification dictates treatment. AAD is classified based on where the dissection of the aorta occurs. If the dissection involves the ascending aorta, it is classified as a type A AAD and should immediately be treated with emergent surgery in order to prevent complications including myocardial infarction, cardiac tamponade, and aortic rupture.^2,4,5 If the dissection is limited to the descending aorta, it is classified as a type B AAD and can be medically managed by controlling pain and lowering blood pressure; if symptoms persist, surgical management may be required.² After hospital discharge, AAD patients are followed closely with medical therapy, serial imaging, and reoperation if necessary.⁴

Our patient underwent emergent surgery for aortic root/ascending aortic replacement with a mechanical valve. He tolerated the procedure well. Surgical tissue pathology of the aortic segment showed a wall of elastic vessel with medial degeneration and dissection, and the tissue pathology of the aorta leaflets showed valvular tissue with myxoid degeneration.

THE TAKEAWAY

It is critical to keep AAD in the differential diagnosis of a patient presenting with acute onset of chest pain, as AAD often has an atypical presentation and can easily be misdiagnosed. Effective imaging is crucial to diagnosis, and immediate treatment is essential to patient survival.

CORRESPONDENCE
Rachel A. Reedy, PA, University of Florida, Department of General Pediatrics, 7046 SW Archer Road, Gainesville, FL 32608; [email protected]

Primary care physicians (PCPs) generate only a small part of the $75 billion to $100 billion wasted every year on low-value care, according to a brief report published online Jan. 18 in Annals of Internal Medicine.

However, one expert said there are better ways to curb low-value care than focusing on which specialties are guilty of the practice.

Analyzing a 20% random sample of Medicare Part B claims, Aaron Baum, PhD, with the Icahn School of Medicine at Mount Sinai, New York, and colleagues found that the services primary care physicians performed or ordered made up on average 8.3% of the low-value care their patients received (interquartile range, 3.9%-15.1%; 95th percentile, 35.6%) and their referrals made up 15.4% (IQR, 6.3%-26.4%; 95th percentile, 44.6%).

By specialty, cardiology had the worst record with 27% of all spending on low-value services ($1.8 billion) attributed to that specialty. Yet, of the 25 highest-spending specialties in the report, 12 of them were associated with 1% or less than 1% each of all low-value spending, indicating the waste was widely distributed.

Dr. Baum said in an interview that though there are some PCPs guilty of high spending on low-value services, overall, most primary care physicians’ low-value services add up to only 0.3% of Part B spending. He noted that Part B spending is about one-third of all Medicare spending.

Primary care is often thought to be at the core of care management and spending and PCPs are often seen as the gatekeepers, but this analysis suggests that efforts to make big differences in curtailing low-value spending might be more effective elsewhere.

“There’s only so much spending you can reduce by changing primary care physicians’ services that they directly perform,” Dr. Baum said.
 

Low-value care is costly, can be harmful

Mark Fendrick, MD, director of the University of Michigan’s Center for Value-Based Insurance Design in Ann Arbor, said in an interview that the report adds confirmation to previous research that has consistently shown low-value care is “extremely common, very costly, and provided by primary care providers and specialists alike.” He noted that it can also be harmful.

“The math is simple,” he said. “If we want to improve coverage and lower patient costs for essential services like visits, diagnostic tests, and drugs, we have to reduce spending on those services that do not make Americans any healthier.”

The study ranked 31 clinical services judged to be low value by physician societies, Medicare and clinical guidelines, and their use among beneficiaries enrolled between 2007 and 2014. Here’s how the top six low-value services compare.

Dr. Fendrick said a weakness of the paper is the years of the data (2007-2014). Some of the criteria around low-value care have changed since then. The age that a prostate-specific antigen test becomes low-value is now 70 years, for instance, instead of 75. He added that some of the figures attributed to non-PCP providers appear out of date.

Dr. Fendrick said, “I understand that there are Medicare patients who end up at a gastroenterologist or surgeon’s office to get colorectal cancer screening, but it would be very hard for me to believe that half of stress tests and over half of colon cancer screening over [age] 85 [years] and half of PSA for people over 75 did not have some type of referring clinicians involved. I certainly don’t think that would be the case in 2020-2021.”

Dr. Baum said those years were the latest years available for the data points needed for this analysis, but he and his colleagues were working to update the data for future publication.

Dr. Fendrick said not much has changed in recent years in terms of waste on low-value care, even with campaigns such as Choosing Wisely dedicated to identifying low-value services or procedures in each specialty.

“I believe there’s not a particular group of clinicians one way or the other who are actually doing any better now than they were 7 years ago,” he said. He would rather focus less on which specialties are associated with the most low-value care and more on the underlying policies that encourage low-value care.

“If you’re going to get paid for doing a stress test and get paid nothing or significantly less if you don’t, the incentives are in the wrong direction,” he said.

Dr. Fendrick said the pandemic era provides an opportunity to eliminate low-value care because use of those services has dropped drastically as resources have been diverted to COVID-19 patients and many services have been delayed or canceled.

He said he has been pushing an approach that providers should be paid more after the pandemic “to do the things we want them to do.”

As an example, he said, instead of paying $886 million on colonoscopies for people over the age of 85, “why don’t we put a policy in place that would make it better for patients by lowering cost sharing and better for providers by paying them more to do the service on the people who need it as opposed to the people who don’t?”

The research was funded by the American Board of Family Medicine Foundation. Dr. Baum and a coauthor reported receiving personal fees from American Board of Family Medicine Foundation during the conduct of the study. Another coauthor reported receiving personal fees from Collective Health, HealthRight 360, PLOS Medicine, and the New England Journal of Medicine, outside the submitted work. Dr. Fendrick disclosed no relevant financial relationships.

A version of this article first appeared on Medscape.com.

Primary care physicians (PCPs) generate only a small part of the $75 billion to $100 billion wasted every year on low-value care, according to a brief report published online Jan. 18 in Annals of Internal Medicine.

However, one expert said there are better ways to curb low-value care than focusing on which specialties are guilty of the practice.

Analyzing a 20% random sample of Medicare Part B claims, Aaron Baum, PhD, with the Icahn School of Medicine at Mount Sinai, New York, and colleagues found that the services primary care physicians performed or ordered made up on average 8.3% of the low-value care their patients received (interquartile range, 3.9%-15.1%; 95th percentile, 35.6%) and their referrals made up 15.4% (IQR, 6.3%-26.4%; 95th percentile, 44.6%).

By specialty, cardiology had the worst record with 27% of all spending on low-value services ($1.8 billion) attributed to that specialty. Yet, of the 25 highest-spending specialties in the report, 12 of them were associated with 1% or less than 1% each of all low-value spending, indicating the waste was widely distributed.

Dr. Baum said in an interview that though there are some PCPs guilty of high spending on low-value services, overall, most primary care physicians’ low-value services add up to only 0.3% of Part B spending. He noted that Part B spending is about one-third of all Medicare spending.

Primary care is often thought to be at the core of care management and spending and PCPs are often seen as the gatekeepers, but this analysis suggests that efforts to make big differences in curtailing low-value spending might be more effective elsewhere.

“There’s only so much spending you can reduce by changing primary care physicians’ services that they directly perform,” Dr. Baum said.
 

Low-value care is costly, can be harmful

Mark Fendrick, MD, director of the University of Michigan’s Center for Value-Based Insurance Design in Ann Arbor, said in an interview that the report adds confirmation to previous research that has consistently shown low-value care is “extremely common, very costly, and provided by primary care providers and specialists alike.” He noted that it can also be harmful.

“The math is simple,” he said. “If we want to improve coverage and lower patient costs for essential services like visits, diagnostic tests, and drugs, we have to reduce spending on those services that do not make Americans any healthier.”

The study ranked 31 clinical services judged to be low value by physician societies, Medicare and clinical guidelines, and their use among beneficiaries enrolled between 2007 and 2014. Here’s how the top six low-value services compare.

Dr. Fendrick said a weakness of the paper is the years of the data (2007-2014). Some of the criteria around low-value care have changed since then. The age that a prostate-specific antigen test becomes low-value is now 70 years, for instance, instead of 75. He added that some of the figures attributed to non-PCP providers appear out of date.

Dr. Fendrick said, “I understand that there are Medicare patients who end up at a gastroenterologist or surgeon’s office to get colorectal cancer screening, but it would be very hard for me to believe that half of stress tests and over half of colon cancer screening over [age] 85 [years] and half of PSA for people over 75 did not have some type of referring clinicians involved. I certainly don’t think that would be the case in 2020-2021.”

Dr. Baum said those years were the latest years available for the data points needed for this analysis, but he and his colleagues were working to update the data for future publication.

Dr. Fendrick said not much has changed in recent years in terms of waste on low-value care, even with campaigns such as Choosing Wisely dedicated to identifying low-value services or procedures in each specialty.

“I believe there’s not a particular group of clinicians one way or the other who are actually doing any better now than they were 7 years ago,” he said. He would rather focus less on which specialties are associated with the most low-value care and more on the underlying policies that encourage low-value care.

“If you’re going to get paid for doing a stress test and get paid nothing or significantly less if you don’t, the incentives are in the wrong direction,” he said.

Dr. Fendrick said the pandemic era provides an opportunity to eliminate low-value care because use of those services has dropped drastically as resources have been diverted to COVID-19 patients and many services have been delayed or canceled.

He said he has been pushing an approach that providers should be paid more after the pandemic “to do the things we want them to do.”

As an example, he said, instead of paying $886 million on colonoscopies for people over the age of 85, “why don’t we put a policy in place that would make it better for patients by lowering cost sharing and better for providers by paying them more to do the service on the people who need it as opposed to the people who don’t?”

The research was funded by the American Board of Family Medicine Foundation. Dr. Baum and a coauthor reported receiving personal fees from American Board of Family Medicine Foundation during the conduct of the study. Another coauthor reported receiving personal fees from Collective Health, HealthRight 360, PLOS Medicine, and the New England Journal of Medicine, outside the submitted work. Dr. Fendrick disclosed no relevant financial relationships.

A version of this article first appeared on Medscape.com.

Primary care physicians (PCPs) generate only a small part of the $75 billion to $100 billion wasted every year on low-value care, according to a brief report published online Jan. 18 in Annals of Internal Medicine.

However, one expert said there are better ways to curb low-value care than focusing on which specialties are guilty of the practice.

Analyzing a 20% random sample of Medicare Part B claims, Aaron Baum, PhD, with the Icahn School of Medicine at Mount Sinai, New York, and colleagues found that the services primary care physicians performed or ordered made up on average 8.3% of the low-value care their patients received (interquartile range, 3.9%-15.1%; 95th percentile, 35.6%) and their referrals made up 15.4% (IQR, 6.3%-26.4%; 95th percentile, 44.6%).

By specialty, cardiology had the worst record with 27% of all spending on low-value services ($1.8 billion) attributed to that specialty. Yet, of the 25 highest-spending specialties in the report, 12 of them were associated with 1% or less than 1% each of all low-value spending, indicating the waste was widely distributed.

Dr. Baum said in an interview that though there are some PCPs guilty of high spending on low-value services, overall, most primary care physicians’ low-value services add up to only 0.3% of Part B spending. He noted that Part B spending is about one-third of all Medicare spending.

Primary care is often thought to be at the core of care management and spending and PCPs are often seen as the gatekeepers, but this analysis suggests that efforts to make big differences in curtailing low-value spending might be more effective elsewhere.

“There’s only so much spending you can reduce by changing primary care physicians’ services that they directly perform,” Dr. Baum said.
 

Low-value care is costly, can be harmful

Mark Fendrick, MD, director of the University of Michigan’s Center for Value-Based Insurance Design in Ann Arbor, said in an interview that the report adds confirmation to previous research that has consistently shown low-value care is “extremely common, very costly, and provided by primary care providers and specialists alike.” He noted that it can also be harmful.

“The math is simple,” he said. “If we want to improve coverage and lower patient costs for essential services like visits, diagnostic tests, and drugs, we have to reduce spending on those services that do not make Americans any healthier.”

The study ranked 31 clinical services judged to be low value by physician societies, Medicare and clinical guidelines, and their use among beneficiaries enrolled between 2007 and 2014. Here’s how the top six low-value services compare.

Dr. Fendrick said a weakness of the paper is the years of the data (2007-2014). Some of the criteria around low-value care have changed since then. The age that a prostate-specific antigen test becomes low-value is now 70 years, for instance, instead of 75. He added that some of the figures attributed to non-PCP providers appear out of date.

Dr. Fendrick said, “I understand that there are Medicare patients who end up at a gastroenterologist or surgeon’s office to get colorectal cancer screening, but it would be very hard for me to believe that half of stress tests and over half of colon cancer screening over [age] 85 [years] and half of PSA for people over 75 did not have some type of referring clinicians involved. I certainly don’t think that would be the case in 2020-2021.”

Dr. Baum said those years were the latest years available for the data points needed for this analysis, but he and his colleagues were working to update the data for future publication.

Dr. Fendrick said not much has changed in recent years in terms of waste on low-value care, even with campaigns such as Choosing Wisely dedicated to identifying low-value services or procedures in each specialty.

“I believe there’s not a particular group of clinicians one way or the other who are actually doing any better now than they were 7 years ago,” he said. He would rather focus less on which specialties are associated with the most low-value care and more on the underlying policies that encourage low-value care.

“If you’re going to get paid for doing a stress test and get paid nothing or significantly less if you don’t, the incentives are in the wrong direction,” he said.

Dr. Fendrick said the pandemic era provides an opportunity to eliminate low-value care because use of those services has dropped drastically as resources have been diverted to COVID-19 patients and many services have been delayed or canceled.

He said he has been pushing an approach that providers should be paid more after the pandemic “to do the things we want them to do.”

As an example, he said, instead of paying $886 million on colonoscopies for people over the age of 85, “why don’t we put a policy in place that would make it better for patients by lowering cost sharing and better for providers by paying them more to do the service on the people who need it as opposed to the people who don’t?”

The research was funded by the American Board of Family Medicine Foundation. Dr. Baum and a coauthor reported receiving personal fees from American Board of Family Medicine Foundation during the conduct of the study. Another coauthor reported receiving personal fees from Collective Health, HealthRight 360, PLOS Medicine, and the New England Journal of Medicine, outside the submitted work. Dr. Fendrick disclosed no relevant financial relationships.

A version of this article first appeared on Medscape.com.

Medication overuse headache (MOH), a secondary headache diagnosis, is a prevalent phenomenon that complicates headache diagnosis and treatment, increases the cost of care, and reduces quality of life. Effective abortive medication is essential for the headache sufferer; when an abortive is used too frequently, however, headache frequency increases—potentially beginning a cycle in which the patient then takes more medication to abort the headache. Over time, the patient suffers from an ever-­increasing number of headaches, takes even more abortive medication, and so on. In the presence of MOH, there is a reduction in pain response to preventive and abortive treatments; when medication overuse is eliminated, pain response improves.¹

Although MOH is well recognized among headache specialists, the condition is often overlooked in primary care. Since headache is a top complaint in primary care, however, and prevention is a major goal in family medicine, the opportunity for you to recognize, treat, and prevent MOH is significant. In fact, a randomized controlled trial showed that brief patient education about headache care and MOH provided by a primary care physician can lead to a significant reduction in headache frequency among patients with MOH.²

Although medication overuse headache is well recognized among headache specialists, the condition is often overlooked in primary care.

This article reviews the recognition and diagnosis of MOH, based on historical features and current criteria; addresses risk factors for abortive medication overuse and how to withdraw an offending agent; and explores the value of bridging and preventive therapies to reduce the overall frequency of headache.

IMAGE © ROY SCOTT

What defines MOH?

Typically, MOH is a chronification of a primary headache disorder. However, in patients with a history of migraine who are undergoing treatment for another chronic pain condition with an opioid or other analgesic, MOH can be induced.³ An increase in the frequency of headache raises the specter of a concomitant increase in the level of disability⁴; psychiatric comorbidity⁵; and more headache days, with time lost from school and work.

The Migraine Disability Assessment (MIDAS) questionnaire, a validated instrument that helps the provider (1) measure the impact that headache has on a patient’s life and (2) follow treatment progress, also provides information to employers and insurance companies on treatment coverage and the need for work modification. The MIDAS score is 3 times higher in patients with MOH than in patients with episodic migraine.^6,7

The annual associated cost per person of MOH has been estimated at $4000, resulting in billions of dollars in associated costs⁸; most of these costs are related to absenteeism and disability. After detoxification for MOH, annual outpatient medication costs are reduced by approximately 24%.⁹

Efforts to solve a common problem create another

Headache affects nearly 50% of the general population worldwide,¹⁰ accounting for about 4% of primary care visits¹¹ and approximately 20% of outpatient neurology consultations.¹² Although inpatient stays for headache are approximately half the duration of the overall average hospital stay, headache accounts for 3% of admissions.¹³ According to the Global Burden of Disease study, tension-type headache, migraine, and MOH are the 3 most common headache disorders.¹⁰ Headache is the second leading cause of disability among people 15 to 49 years of age.¹⁰

Continue to: The prevalence of MOH...

The prevalence of MOH in the general population is 2%.^7,14,15 A population-based study showed that the rate of progression from episodic headache (< 15 d/mo) to chronic headache (≥ 15 d/mo) in the general population is 2.5% per year¹⁶; however, progression to chronic headache is 14% per year in patients with medication overuse. One-third of the general population with chronic migraine overuses symptomatic medication; in US headache clinics, roughly one-half of patients with chronic headache overuse acute medication.⁶

Definitions and diagnosis

MOH is a secondary headache diagnosis in the third edition of the International Classification of Headache Disorders (ICHD-3) (TABLE 1),¹⁷ which lists diagnostic criteria for recognized headache disorders.

Terminology. MOH has also been called rebound headache, drug-induced headache, and transformed migraine, but these terms are outdated and are not formal diagnoses. Patients sometimes refer to substance-withdrawal headaches (not discussed in this article) as rebound headaches, so clarity is important when discussing headache with patients: namely, that MOH is an exacerbation of an existing headache condition caused by overuse of abortive headache medications, including analgesics, combination analgesics, triptans, barbiturates, and opioids.

The time it takes to develop medication overuse headache is shortest with triptans, followed by ergots, then analgesics.

MOH was recognized in the early 1950s and fully differentiated as a diagnosis in 2005 in the second edition of the ICHD. The disorder is subcategorized by offending abortive agent (TABLE 2¹⁷) because the frequency of analgesic use required to develop MOH differs by agent.

Risk factors for MOH and chronification of a primary headache ­disorder. There are several risk factors for developing MOH, and others that contribute to increasing headache frequency in general (TABLE 3^5,14,18-23). Some risk factors are common to each. All are important to address because some are modifiable.

Continue to: Pathophysiology

Pathophysiology. The pathophysiology and psychology behind MOH are largely unknown. Physiologic changes in pain processing and functional imaging changes have been demonstrated in patients with MOH, both of which are reversible upon withdrawal of medication.²³ Genetic factors and changes in hormone and neurotransmitter levels are found in MOH patients; this is not the case in patients who have an episodic headache pattern only.²⁴

Presentation. Diagnostic criteria for MOH do not include clinical characteristics. Typically, the phenotype of MOH in a given patient is similar to the underlying primary headache²⁵—although this principle can be complicated to tease out because these medications can suppress some symptoms. Diagnosis of a primary headache disorder should be documented along with the diagnosis of MOH.

Medication overuse can exist without MOH: Not every patient who frequently uses an abortive medication develops MOH.

Treatment is multifaceted—and can become complex

Mainstays of treatment of MOH are education about the disorder and detoxification from the overused agent, although specific treatments can differ depending on the agent involved, the frequency and duration of its use, and a patient’s behavioral patterns and psychiatric comorbidities. Often, a daily medication to prevent headache is considered upon, or after, withdrawal of the offending agent. The timing of introducing a preventive might impact its effectiveness. Some refractory cases require more intensive therapy, including hospitalization at a specialized tertiary center.

But before we look at detoxification from an overused agent, it’s important to review one of the best strategies of all in combatting MOH.

Continue to: First and best strategy

Select an appropriate abortive to reduce the risk of MOH. With regard to specific acute headache medications, some nuances other than type of headache should be considered. Nonsteroidal anti-inflammatory drugs (NSAIDs) are recommended as abortive therapy by the American Headache Society for their efficacy, favorable adverse effect profile, and low cost. NSAIDs are protective against development of MOH if a patient’s baseline headache frequency is < 10/mo; at a frequency of 10 to 14 d/mo, however, the risk of MOH increases when using an NSAID.⁶ A similar effect has been seen with triptans.¹⁶ Longer-acting NSAIDs, such as nabumetone and naproxen, have been proposed as less likely to cause MOH, and are even used as bridging therapy sometimes (as long as neither of these was the overused medication).²⁶

The time it takes to develop MOH is shortest with triptans, followed by ergots, then analgesics.²⁷

Prospective cohort studies^6,16 have shown that barbiturates and opioids are more likely to induce MOH; for that reason, agents in these analgesic classes are almost universally avoided unless no other medically acceptable options exist. Using barbiturate-containing compounds or opioids > 4 d/mo exponentially increases the likelihood of MOH.

Promising preclinical data demonstrate that the gepant, or small-molecule calcitonin gene-related peptide (CGRP) receptor antagonist, class of medications used as abortive therapy does not induce medication overuse cutaneous allodynia.²⁸

Provide education. Primary prevention of MOH involves (1) increasing patients’ awareness of how to take medications appropriately and (2) restricting intake of over-the-counter abortive medications. Often, the expert recommendation is to limit abortives to approximately 2 d/wk because more frequent use places patients at risk of further increased use and subsequent MOH.

Continue to: A randomized controlled trial in Norway...

A randomized controlled trial in Norway compared outcomes in 2 groups of patients with MOH: One group was given advice on the disorder by a physician; the other group was not provided with advice. In the “business-as-usual” group, there was no significant improvement; however, when general practitioners provided simple advice (lasting roughly 9 minutes) about reducing abortive medication use to a safe level and cautioned patients that they would be “feeling worse before feeling better,” headache days were reduced by approximately 8 per month and medication days, by 16 per month.²

A subsequent, long-term follow-up study²⁹ of patients from the Norway trial² who had been given advice and education showed a relapse rate (ie, into overuse of headache medication) of only 8% and sustained reduction of headache days and medication use at 16 months.

Offer support and other nondrug interventions. A recent review of 3 studies²³ recommended that extra support for patients from a headache nurse, close follow-up, keeping an electronic diary that provides feedback, and undertaking a short course of psychotherapy can reduce medication overuse and prevent relapse.

If MOH develops, initiate withdrawal, introduce a preventive

Withdraw overused medication. Most current evidence suggests that withdrawal of the offending agent is the most effective factor in reducing headache days and improving quality of life. A randomized controlled trial compared the effects of (1) complete and immediate withdrawal of an abortive medication with (2) reducing its use (ie, limiting intake to 2 d/wk), on headache frequency, disability, and quality of life.³⁰ There was a reduction of headache days in both groups; however, reduction was much greater at 2 months in the complete withdrawal group than in the restricted intake group (respectively, a 41% and a 26% reduction in headache days per month). This effect was sustained at 6 and 12 months in both groups. The study confirmed the results of earlier research^2,15: Abrupt withdrawal leads to reversion to an episodic pattern at 2 to 6 months in approximately 40% to 60% of patients.

More studies are needed to determine the most appropriate treatment course for MOH; however, complete withdrawal of the causative drug is the most important intervention.

Continue to: Consider withdrawal plus preventive treatment

Consider withdrawal plus preventive treatment. Use of sodium valproate, in addition to medication overuse detoxification, led to a significant reduction in headache days and improvement in quality of life at 12 weeks but no difference after 24 weeks, compared with detoxification alone in a randomized, double-blind, placebo-controlled study.³¹

A study of 61 patients showed a larger reduction (by 7.2 d/mo) in headache frequency with any preventive medication in addition to medication withdrawal, compared to withdrawal alone (by 4.1 d/mo) after 3 months; however, the relative benefit was gone at 6 months.³²

A study of 98 patients compared immediate and delayed initiation of preventive medication upon withdrawal of overused abortive medication.³³ Response was defined as a > 50% reduction in headache frequency and was similar in both groups; results showed a 28% response with immediate initiation of a preventive; a 23% response with delayed (ie, 2 months after withdrawal) initiation; and a 48% response in both groups at 12 months.

Collectively, these studies suggest that adding a preventive medication at the time of withdrawal has the potential to reduce headache frequency more quickly than withdrawal alone. However, after 3 to 6 months, the outcome of reduced headache frequency is the same whether or not a preventive medication is used—as long as the offending agent has been withdrawn.

Do preventives work without withdrawing overused medication? Patients with MOH often show little or no improvement with addition of a preventive medication only; their response to a preventive improves after withdrawal of the overused medication. Patients without previous headache improvement after addition of a preventive, who also did not improve 2 months after withdrawal, then demonstrated an overall reduction in headache by 26% when a preventive was reintroduced after withdrawal.²

Continue to: The research evidence for preventives

The research evidence for preventives. Medications for headache prevention have not been extensively evaluated specifically for treating MOH. Here is what’s known:

• Flunarizine, amitriptyline, and beta-blockers usually are ineffective for MOH.²⁴
• Results for topiramate are mixed: A small, double-blind, placebo-controlled chronic migraine study in Europe showed that, in a subgroup of patients with MOH, topiramate led to a small but significant reduction (3.5 d/mo) in headache frequency, compared to placebo.²⁷ A similar study done in the United States did not show a significant difference between the active-treatment and placebo groups.³⁴
• Findings regarding onabotulinumtoxinA are intriguing: In a posthoc analysis of onabotulinumtoxinA to treat chronic migraine, patients with MOH who did not undergo detoxification had an 8 d/mo greater reduction in headache, compared to placebo.³⁵ However, when compared to placebo in conjunction with detoxification, onabotulinumtoxinA demonstrated no benefit.³⁶
• Newer CGRP antagonist and CGRP receptor antagonist monoclonal antibodies are successful preventive medications that have demonstrated a reduction in acute medication use days per month and headache days per month³⁷; these compounds have not been compared to withdrawal alone.

Reducing the severity and duration of withdrawal symptoms

Withdrawal from overused abortive headache medications can lead to worsening headache, nausea, vomiting, hypotension, tachycardia, sleep disturbances, restlessness, anxiety, and nervousness. Symptoms usually last 2 to 10 days but can persist for as long as 4 weeks; duration of withdrawal symptoms varies with the medication that is being overused. In patients who have used a triptan, for example, mean duration of withdrawal is 4.1 days; ergotamine, 6.7 days; and NSAIDs, 9.5 days.²³ Tapered withdrawal is sometimes recommended with opioids and barbiturates to reduce withdrawal symptoms. It is unclear whether starting a preventive medication during withdrawal assists in reducing withdrawal symptoms.³⁸

Bridging therapy to reduce symptoms of withdrawal is often provided despite debatable utility. Available evidence does not favor one agent or method but suggests some strategies that could be helpful:

• A prednisone taper has a potential role during the first 6 days of withdrawal by reducing rebound headache and withdrawal symptoms³⁹; however, oral prednisolone has been shown to have no benefit.⁴⁰
• Alone, IV methylprednisolone seems not to be of benefit; however, in a retrospective study of 94 patients, IV methylprednisolone plus diazepam for 5 days led to a significant reduction in headache frequency and drug consumption that was sustained after 3 months.⁴¹
• Celecoxib was compared to prednisone over a 20-day course: a celecoxib dosage of 400 mg/d for the first 5 days, tapered by 100 mg every 5 days, and an oral prednisone dosage of 75 mg/d for the first 5 days, then tapered every 5 days. Patients taking celecoxib had lower headache intensity but there was no difference in headache frequency and acute medication intake between the groups.⁴²

Other strategies. Using antiemetics and NSAIDs to reduce withdrawal symptoms is widely practiced, but no placebo-­controlled trials have been conducted to support this strategy.

Reduce the risk of medication overuse headache by selecting an appropriate abortive; NSAIDs are recommended for their efficacy, favorable adverse effect profile, and low cost.

Patients in withdrawal might be more likely to benefit from inpatient care if they have a severe comorbidity, such as opioid or barbiturate use; failure to respond to, tolerate, or adhere to treatment; or relapse after withdrawal.³⁸

Continue to: Cognitive behavioral therapy...

Cognitive behavioral therapy, exercise, a headache diary, and biofeedback should be considered in every patient’s treatment strategy because a multidisciplinary approach increases adherence and leads to improvement in headache frequency and a decrease in disability and medication use.⁴³

Predictors of Tx success

A prospective cohort study determined that the rate of MOH relapse is 31% at 6 months, 41% at 1 year, and 45% at 4 years, with the highest risk of relapse during the first year.⁴⁴ Looking at the correlation between type of medication overused and relapse rate, the research indicates that

• triptans have the lowest risk of relapse,⁴⁴
• simple analgesics have a higher risk of relapse than triptans,^22,44 and
• opioids have the highest risk of relapse.²²

Where the data don’t agree. Data on combination analgesics and on ergots are conflicting.²² In addition, data on whether the primary type of headache predicts relapse rate conflict; however, migraine might predict a better outcome than tension-type headache.²²

To recap and expand: Management pearls

The major goals of headache management generally are to rule out secondary headache, reach a correct diagnosis, reduce overall headache frequency, and provide effective abortive medication. A large component of reducing headache frequency is addressing and treating medication overuse.

Seek to understand the nature of the patient’s headache disorder. Components of the history are key in identifying the underlying headache diagnosis and ruling out other, more concerning secondary headache diagnoses. The ICHD-3 is an excellent resource for treating headache disorders because the classification lists specific diagnostic criteria for all recognized headache diagnoses.

Continue to: Medication withdrawal...

Medication withdrawal—with or without preventive medication—should reduce the frequency of MOH in 2 or 3 months. If headache does not become less frequent, however, the headache diagnosis might need to be reconsidered. Minimizing the use of abortive medication is generally recommended, but reduction or withdrawal of these medications does not guarantee that patients will revert to an episodic pattern of headache.

Inpatient care of withdrawal might be beneficial when a patient has a severe comorbidity; does not respond to, tolerate, or adhere to treatment; or relapses after withdrawal.

Treating withdrawal symptoms is a reasonable approach in some patients, but evidence does not support routinely providing bridging therapy.

Apply preventives carefully. Abortive medication withdrawal should generally be completed before initiating preventive medication; however, over the short term, starting preventive therapy while withdrawing the overused medication could assist in reducing headache frequency rapidly. This strategy can put patients at risk of medication adverse effects and using the medications longer than necessary, yet might be reasonable in certain patients, given their comorbidities, risk of relapse, and physician and patient preference. A preventive medication for an individual patient should generally be chosen in line with recommendations of the American Academy of Neurology⁴⁵ and on the basis of the history and comorbidities.

Provide education, which is essential to lowering barriers to success. Patients with MOH must be counseled to understand that (1) a headache treatment that is supposed to be making them feel better is, in fact, making them feel worse and (2) they will get worse before they get better. Many patients are afraid to be without medication to use as needed. It is helpful to educate them on the different types of treatments (abortive, preventive); how MOH interferes with headache prophylaxis and medication efficacy; how MOH alters brain function (ie, aforementioned physiologic changes in pain processing and functional imaging changes²³); and that such change is reversible when medication is withdrawn.

ACKNOWLEDGEMENT
The author thanks Jeffrey Curtis, MD, MPH, for his support and editing assistance with the manuscript.

CORRESPONDENCE
Allison Crain, MD, 2927 N 7th Avenue, Phoenix, AZ 85013; [email protected].

Medication overuse headache (MOH), a secondary headache diagnosis, is a prevalent phenomenon that complicates headache diagnosis and treatment, increases the cost of care, and reduces quality of life. Effective abortive medication is essential for the headache sufferer; when an abortive is used too frequently, however, headache frequency increases—potentially beginning a cycle in which the patient then takes more medication to abort the headache. Over time, the patient suffers from an ever-­increasing number of headaches, takes even more abortive medication, and so on. In the presence of MOH, there is a reduction in pain response to preventive and abortive treatments; when medication overuse is eliminated, pain response improves.¹

Although MOH is well recognized among headache specialists, the condition is often overlooked in primary care. Since headache is a top complaint in primary care, however, and prevention is a major goal in family medicine, the opportunity for you to recognize, treat, and prevent MOH is significant. In fact, a randomized controlled trial showed that brief patient education about headache care and MOH provided by a primary care physician can lead to a significant reduction in headache frequency among patients with MOH.²

Although medication overuse headache is well recognized among headache specialists, the condition is often overlooked in primary care.

This article reviews the recognition and diagnosis of MOH, based on historical features and current criteria; addresses risk factors for abortive medication overuse and how to withdraw an offending agent; and explores the value of bridging and preventive therapies to reduce the overall frequency of headache.

IMAGE © ROY SCOTT

What defines MOH?

Typically, MOH is a chronification of a primary headache disorder. However, in patients with a history of migraine who are undergoing treatment for another chronic pain condition with an opioid or other analgesic, MOH can be induced.³ An increase in the frequency of headache raises the specter of a concomitant increase in the level of disability⁴; psychiatric comorbidity⁵; and more headache days, with time lost from school and work.

The Migraine Disability Assessment (MIDAS) questionnaire, a validated instrument that helps the provider (1) measure the impact that headache has on a patient’s life and (2) follow treatment progress, also provides information to employers and insurance companies on treatment coverage and the need for work modification. The MIDAS score is 3 times higher in patients with MOH than in patients with episodic migraine.^6,7

The annual associated cost per person of MOH has been estimated at $4000, resulting in billions of dollars in associated costs⁸; most of these costs are related to absenteeism and disability. After detoxification for MOH, annual outpatient medication costs are reduced by approximately 24%.⁹

Efforts to solve a common problem create another

Headache affects nearly 50% of the general population worldwide,¹⁰ accounting for about 4% of primary care visits¹¹ and approximately 20% of outpatient neurology consultations.¹² Although inpatient stays for headache are approximately half the duration of the overall average hospital stay, headache accounts for 3% of admissions.¹³ According to the Global Burden of Disease study, tension-type headache, migraine, and MOH are the 3 most common headache disorders.¹⁰ Headache is the second leading cause of disability among people 15 to 49 years of age.¹⁰

Continue to: The prevalence of MOH...

The prevalence of MOH in the general population is 2%.^7,14,15 A population-based study showed that the rate of progression from episodic headache (< 15 d/mo) to chronic headache (≥ 15 d/mo) in the general population is 2.5% per year¹⁶; however, progression to chronic headache is 14% per year in patients with medication overuse. One-third of the general population with chronic migraine overuses symptomatic medication; in US headache clinics, roughly one-half of patients with chronic headache overuse acute medication.⁶

Definitions and diagnosis

MOH is a secondary headache diagnosis in the third edition of the International Classification of Headache Disorders (ICHD-3) (TABLE 1),¹⁷ which lists diagnostic criteria for recognized headache disorders.

Terminology. MOH has also been called rebound headache, drug-induced headache, and transformed migraine, but these terms are outdated and are not formal diagnoses. Patients sometimes refer to substance-withdrawal headaches (not discussed in this article) as rebound headaches, so clarity is important when discussing headache with patients: namely, that MOH is an exacerbation of an existing headache condition caused by overuse of abortive headache medications, including analgesics, combination analgesics, triptans, barbiturates, and opioids.

The time it takes to develop medication overuse headache is shortest with triptans, followed by ergots, then analgesics.

MOH was recognized in the early 1950s and fully differentiated as a diagnosis in 2005 in the second edition of the ICHD. The disorder is subcategorized by offending abortive agent (TABLE 2¹⁷) because the frequency of analgesic use required to develop MOH differs by agent.

Risk factors for MOH and chronification of a primary headache ­disorder. There are several risk factors for developing MOH, and others that contribute to increasing headache frequency in general (TABLE 3^5,14,18-23). Some risk factors are common to each. All are important to address because some are modifiable.

Continue to: Pathophysiology

Pathophysiology. The pathophysiology and psychology behind MOH are largely unknown. Physiologic changes in pain processing and functional imaging changes have been demonstrated in patients with MOH, both of which are reversible upon withdrawal of medication.²³ Genetic factors and changes in hormone and neurotransmitter levels are found in MOH patients; this is not the case in patients who have an episodic headache pattern only.²⁴

Presentation. Diagnostic criteria for MOH do not include clinical characteristics. Typically, the phenotype of MOH in a given patient is similar to the underlying primary headache²⁵—although this principle can be complicated to tease out because these medications can suppress some symptoms. Diagnosis of a primary headache disorder should be documented along with the diagnosis of MOH.

Medication overuse can exist without MOH: Not every patient who frequently uses an abortive medication develops MOH.

Treatment is multifaceted—and can become complex

Mainstays of treatment of MOH are education about the disorder and detoxification from the overused agent, although specific treatments can differ depending on the agent involved, the frequency and duration of its use, and a patient’s behavioral patterns and psychiatric comorbidities. Often, a daily medication to prevent headache is considered upon, or after, withdrawal of the offending agent. The timing of introducing a preventive might impact its effectiveness. Some refractory cases require more intensive therapy, including hospitalization at a specialized tertiary center.

But before we look at detoxification from an overused agent, it’s important to review one of the best strategies of all in combatting MOH.

Continue to: First and best strategy

Select an appropriate abortive to reduce the risk of MOH. With regard to specific acute headache medications, some nuances other than type of headache should be considered. Nonsteroidal anti-inflammatory drugs (NSAIDs) are recommended as abortive therapy by the American Headache Society for their efficacy, favorable adverse effect profile, and low cost. NSAIDs are protective against development of MOH if a patient’s baseline headache frequency is < 10/mo; at a frequency of 10 to 14 d/mo, however, the risk of MOH increases when using an NSAID.⁶ A similar effect has been seen with triptans.¹⁶ Longer-acting NSAIDs, such as nabumetone and naproxen, have been proposed as less likely to cause MOH, and are even used as bridging therapy sometimes (as long as neither of these was the overused medication).²⁶

The time it takes to develop MOH is shortest with triptans, followed by ergots, then analgesics.²⁷

Prospective cohort studies^6,16 have shown that barbiturates and opioids are more likely to induce MOH; for that reason, agents in these analgesic classes are almost universally avoided unless no other medically acceptable options exist. Using barbiturate-containing compounds or opioids > 4 d/mo exponentially increases the likelihood of MOH.

Promising preclinical data demonstrate that the gepant, or small-molecule calcitonin gene-related peptide (CGRP) receptor antagonist, class of medications used as abortive therapy does not induce medication overuse cutaneous allodynia.²⁸

Provide education. Primary prevention of MOH involves (1) increasing patients’ awareness of how to take medications appropriately and (2) restricting intake of over-the-counter abortive medications. Often, the expert recommendation is to limit abortives to approximately 2 d/wk because more frequent use places patients at risk of further increased use and subsequent MOH.

Continue to: A randomized controlled trial in Norway...

A randomized controlled trial in Norway compared outcomes in 2 groups of patients with MOH: One group was given advice on the disorder by a physician; the other group was not provided with advice. In the “business-as-usual” group, there was no significant improvement; however, when general practitioners provided simple advice (lasting roughly 9 minutes) about reducing abortive medication use to a safe level and cautioned patients that they would be “feeling worse before feeling better,” headache days were reduced by approximately 8 per month and medication days, by 16 per month.²

A subsequent, long-term follow-up study²⁹ of patients from the Norway trial² who had been given advice and education showed a relapse rate (ie, into overuse of headache medication) of only 8% and sustained reduction of headache days and medication use at 16 months.

Offer support and other nondrug interventions. A recent review of 3 studies²³ recommended that extra support for patients from a headache nurse, close follow-up, keeping an electronic diary that provides feedback, and undertaking a short course of psychotherapy can reduce medication overuse and prevent relapse.

If MOH develops, initiate withdrawal, introduce a preventive

Withdraw overused medication. Most current evidence suggests that withdrawal of the offending agent is the most effective factor in reducing headache days and improving quality of life. A randomized controlled trial compared the effects of (1) complete and immediate withdrawal of an abortive medication with (2) reducing its use (ie, limiting intake to 2 d/wk), on headache frequency, disability, and quality of life.³⁰ There was a reduction of headache days in both groups; however, reduction was much greater at 2 months in the complete withdrawal group than in the restricted intake group (respectively, a 41% and a 26% reduction in headache days per month). This effect was sustained at 6 and 12 months in both groups. The study confirmed the results of earlier research^2,15: Abrupt withdrawal leads to reversion to an episodic pattern at 2 to 6 months in approximately 40% to 60% of patients.

More studies are needed to determine the most appropriate treatment course for MOH; however, complete withdrawal of the causative drug is the most important intervention.

Continue to: Consider withdrawal plus preventive treatment

Consider withdrawal plus preventive treatment. Use of sodium valproate, in addition to medication overuse detoxification, led to a significant reduction in headache days and improvement in quality of life at 12 weeks but no difference after 24 weeks, compared with detoxification alone in a randomized, double-blind, placebo-controlled study.³¹

A study of 61 patients showed a larger reduction (by 7.2 d/mo) in headache frequency with any preventive medication in addition to medication withdrawal, compared to withdrawal alone (by 4.1 d/mo) after 3 months; however, the relative benefit was gone at 6 months.³²

A study of 98 patients compared immediate and delayed initiation of preventive medication upon withdrawal of overused abortive medication.³³ Response was defined as a > 50% reduction in headache frequency and was similar in both groups; results showed a 28% response with immediate initiation of a preventive; a 23% response with delayed (ie, 2 months after withdrawal) initiation; and a 48% response in both groups at 12 months.

Collectively, these studies suggest that adding a preventive medication at the time of withdrawal has the potential to reduce headache frequency more quickly than withdrawal alone. However, after 3 to 6 months, the outcome of reduced headache frequency is the same whether or not a preventive medication is used—as long as the offending agent has been withdrawn.

Do preventives work without withdrawing overused medication? Patients with MOH often show little or no improvement with addition of a preventive medication only; their response to a preventive improves after withdrawal of the overused medication. Patients without previous headache improvement after addition of a preventive, who also did not improve 2 months after withdrawal, then demonstrated an overall reduction in headache by 26% when a preventive was reintroduced after withdrawal.²

Continue to: The research evidence for preventives

The research evidence for preventives. Medications for headache prevention have not been extensively evaluated specifically for treating MOH. Here is what’s known:

• Flunarizine, amitriptyline, and beta-blockers usually are ineffective for MOH.²⁴
• Results for topiramate are mixed: A small, double-blind, placebo-controlled chronic migraine study in Europe showed that, in a subgroup of patients with MOH, topiramate led to a small but significant reduction (3.5 d/mo) in headache frequency, compared to placebo.²⁷ A similar study done in the United States did not show a significant difference between the active-treatment and placebo groups.³⁴
• Findings regarding onabotulinumtoxinA are intriguing: In a posthoc analysis of onabotulinumtoxinA to treat chronic migraine, patients with MOH who did not undergo detoxification had an 8 d/mo greater reduction in headache, compared to placebo.³⁵ However, when compared to placebo in conjunction with detoxification, onabotulinumtoxinA demonstrated no benefit.³⁶
• Newer CGRP antagonist and CGRP receptor antagonist monoclonal antibodies are successful preventive medications that have demonstrated a reduction in acute medication use days per month and headache days per month³⁷; these compounds have not been compared to withdrawal alone.

Reducing the severity and duration of withdrawal symptoms

Withdrawal from overused abortive headache medications can lead to worsening headache, nausea, vomiting, hypotension, tachycardia, sleep disturbances, restlessness, anxiety, and nervousness. Symptoms usually last 2 to 10 days but can persist for as long as 4 weeks; duration of withdrawal symptoms varies with the medication that is being overused. In patients who have used a triptan, for example, mean duration of withdrawal is 4.1 days; ergotamine, 6.7 days; and NSAIDs, 9.5 days.²³ Tapered withdrawal is sometimes recommended with opioids and barbiturates to reduce withdrawal symptoms. It is unclear whether starting a preventive medication during withdrawal assists in reducing withdrawal symptoms.³⁸

Bridging therapy to reduce symptoms of withdrawal is often provided despite debatable utility. Available evidence does not favor one agent or method but suggests some strategies that could be helpful:

• A prednisone taper has a potential role during the first 6 days of withdrawal by reducing rebound headache and withdrawal symptoms³⁹; however, oral prednisolone has been shown to have no benefit.⁴⁰
• Alone, IV methylprednisolone seems not to be of benefit; however, in a retrospective study of 94 patients, IV methylprednisolone plus diazepam for 5 days led to a significant reduction in headache frequency and drug consumption that was sustained after 3 months.⁴¹
• Celecoxib was compared to prednisone over a 20-day course: a celecoxib dosage of 400 mg/d for the first 5 days, tapered by 100 mg every 5 days, and an oral prednisone dosage of 75 mg/d for the first 5 days, then tapered every 5 days. Patients taking celecoxib had lower headache intensity but there was no difference in headache frequency and acute medication intake between the groups.⁴²

Other strategies. Using antiemetics and NSAIDs to reduce withdrawal symptoms is widely practiced, but no placebo-­controlled trials have been conducted to support this strategy.

Reduce the risk of medication overuse headache by selecting an appropriate abortive; NSAIDs are recommended for their efficacy, favorable adverse effect profile, and low cost.

Patients in withdrawal might be more likely to benefit from inpatient care if they have a severe comorbidity, such as opioid or barbiturate use; failure to respond to, tolerate, or adhere to treatment; or relapse after withdrawal.³⁸

Continue to: Cognitive behavioral therapy...

Cognitive behavioral therapy, exercise, a headache diary, and biofeedback should be considered in every patient’s treatment strategy because a multidisciplinary approach increases adherence and leads to improvement in headache frequency and a decrease in disability and medication use.⁴³

Predictors of Tx success

A prospective cohort study determined that the rate of MOH relapse is 31% at 6 months, 41% at 1 year, and 45% at 4 years, with the highest risk of relapse during the first year.⁴⁴ Looking at the correlation between type of medication overused and relapse rate, the research indicates that

• triptans have the lowest risk of relapse,⁴⁴
• simple analgesics have a higher risk of relapse than triptans,^22,44 and
• opioids have the highest risk of relapse.²²

Where the data don’t agree. Data on combination analgesics and on ergots are conflicting.²² In addition, data on whether the primary type of headache predicts relapse rate conflict; however, migraine might predict a better outcome than tension-type headache.²²

To recap and expand: Management pearls

The major goals of headache management generally are to rule out secondary headache, reach a correct diagnosis, reduce overall headache frequency, and provide effective abortive medication. A large component of reducing headache frequency is addressing and treating medication overuse.

Seek to understand the nature of the patient’s headache disorder. Components of the history are key in identifying the underlying headache diagnosis and ruling out other, more concerning secondary headache diagnoses. The ICHD-3 is an excellent resource for treating headache disorders because the classification lists specific diagnostic criteria for all recognized headache diagnoses.

Continue to: Medication withdrawal...

Medication withdrawal—with or without preventive medication—should reduce the frequency of MOH in 2 or 3 months. If headache does not become less frequent, however, the headache diagnosis might need to be reconsidered. Minimizing the use of abortive medication is generally recommended, but reduction or withdrawal of these medications does not guarantee that patients will revert to an episodic pattern of headache.

Inpatient care of withdrawal might be beneficial when a patient has a severe comorbidity; does not respond to, tolerate, or adhere to treatment; or relapses after withdrawal.

Treating withdrawal symptoms is a reasonable approach in some patients, but evidence does not support routinely providing bridging therapy.

Apply preventives carefully. Abortive medication withdrawal should generally be completed before initiating preventive medication; however, over the short term, starting preventive therapy while withdrawing the overused medication could assist in reducing headache frequency rapidly. This strategy can put patients at risk of medication adverse effects and using the medications longer than necessary, yet might be reasonable in certain patients, given their comorbidities, risk of relapse, and physician and patient preference. A preventive medication for an individual patient should generally be chosen in line with recommendations of the American Academy of Neurology⁴⁵ and on the basis of the history and comorbidities.

Provide education, which is essential to lowering barriers to success. Patients with MOH must be counseled to understand that (1) a headache treatment that is supposed to be making them feel better is, in fact, making them feel worse and (2) they will get worse before they get better. Many patients are afraid to be without medication to use as needed. It is helpful to educate them on the different types of treatments (abortive, preventive); how MOH interferes with headache prophylaxis and medication efficacy; how MOH alters brain function (ie, aforementioned physiologic changes in pain processing and functional imaging changes²³); and that such change is reversible when medication is withdrawn.

ACKNOWLEDGEMENT
The author thanks Jeffrey Curtis, MD, MPH, for his support and editing assistance with the manuscript.

CORRESPONDENCE
Allison Crain, MD, 2927 N 7th Avenue, Phoenix, AZ 85013; [email protected].

Medication overuse headache (MOH), a secondary headache diagnosis, is a prevalent phenomenon that complicates headache diagnosis and treatment, increases the cost of care, and reduces quality of life. Effective abortive medication is essential for the headache sufferer; when an abortive is used too frequently, however, headache frequency increases—potentially beginning a cycle in which the patient then takes more medication to abort the headache. Over time, the patient suffers from an ever-­increasing number of headaches, takes even more abortive medication, and so on. In the presence of MOH, there is a reduction in pain response to preventive and abortive treatments; when medication overuse is eliminated, pain response improves.¹

Although MOH is well recognized among headache specialists, the condition is often overlooked in primary care. Since headache is a top complaint in primary care, however, and prevention is a major goal in family medicine, the opportunity for you to recognize, treat, and prevent MOH is significant. In fact, a randomized controlled trial showed that brief patient education about headache care and MOH provided by a primary care physician can lead to a significant reduction in headache frequency among patients with MOH.²

Although medication overuse headache is well recognized among headache specialists, the condition is often overlooked in primary care.

This article reviews the recognition and diagnosis of MOH, based on historical features and current criteria; addresses risk factors for abortive medication overuse and how to withdraw an offending agent; and explores the value of bridging and preventive therapies to reduce the overall frequency of headache.

IMAGE © ROY SCOTT

What defines MOH?

Typically, MOH is a chronification of a primary headache disorder. However, in patients with a history of migraine who are undergoing treatment for another chronic pain condition with an opioid or other analgesic, MOH can be induced.³ An increase in the frequency of headache raises the specter of a concomitant increase in the level of disability⁴; psychiatric comorbidity⁵; and more headache days, with time lost from school and work.

The Migraine Disability Assessment (MIDAS) questionnaire, a validated instrument that helps the provider (1) measure the impact that headache has on a patient’s life and (2) follow treatment progress, also provides information to employers and insurance companies on treatment coverage and the need for work modification. The MIDAS score is 3 times higher in patients with MOH than in patients with episodic migraine.^6,7

The annual associated cost per person of MOH has been estimated at $4000, resulting in billions of dollars in associated costs⁸; most of these costs are related to absenteeism and disability. After detoxification for MOH, annual outpatient medication costs are reduced by approximately 24%.⁹

Efforts to solve a common problem create another

Headache affects nearly 50% of the general population worldwide,¹⁰ accounting for about 4% of primary care visits¹¹ and approximately 20% of outpatient neurology consultations.¹² Although inpatient stays for headache are approximately half the duration of the overall average hospital stay, headache accounts for 3% of admissions.¹³ According to the Global Burden of Disease study, tension-type headache, migraine, and MOH are the 3 most common headache disorders.¹⁰ Headache is the second leading cause of disability among people 15 to 49 years of age.¹⁰

Continue to: The prevalence of MOH...

The prevalence of MOH in the general population is 2%.^7,14,15 A population-based study showed that the rate of progression from episodic headache (< 15 d/mo) to chronic headache (≥ 15 d/mo) in the general population is 2.5% per year¹⁶; however, progression to chronic headache is 14% per year in patients with medication overuse. One-third of the general population with chronic migraine overuses symptomatic medication; in US headache clinics, roughly one-half of patients with chronic headache overuse acute medication.⁶

Definitions and diagnosis

MOH is a secondary headache diagnosis in the third edition of the International Classification of Headache Disorders (ICHD-3) (TABLE 1),¹⁷ which lists diagnostic criteria for recognized headache disorders.

Terminology. MOH has also been called rebound headache, drug-induced headache, and transformed migraine, but these terms are outdated and are not formal diagnoses. Patients sometimes refer to substance-withdrawal headaches (not discussed in this article) as rebound headaches, so clarity is important when discussing headache with patients: namely, that MOH is an exacerbation of an existing headache condition caused by overuse of abortive headache medications, including analgesics, combination analgesics, triptans, barbiturates, and opioids.

The time it takes to develop medication overuse headache is shortest with triptans, followed by ergots, then analgesics.

MOH was recognized in the early 1950s and fully differentiated as a diagnosis in 2005 in the second edition of the ICHD. The disorder is subcategorized by offending abortive agent (TABLE 2¹⁷) because the frequency of analgesic use required to develop MOH differs by agent.

Risk factors for MOH and chronification of a primary headache ­disorder. There are several risk factors for developing MOH, and others that contribute to increasing headache frequency in general (TABLE 3^5,14,18-23). Some risk factors are common to each. All are important to address because some are modifiable.

Continue to: Pathophysiology

Pathophysiology. The pathophysiology and psychology behind MOH are largely unknown. Physiologic changes in pain processing and functional imaging changes have been demonstrated in patients with MOH, both of which are reversible upon withdrawal of medication.²³ Genetic factors and changes in hormone and neurotransmitter levels are found in MOH patients; this is not the case in patients who have an episodic headache pattern only.²⁴

Presentation. Diagnostic criteria for MOH do not include clinical characteristics. Typically, the phenotype of MOH in a given patient is similar to the underlying primary headache²⁵—although this principle can be complicated to tease out because these medications can suppress some symptoms. Diagnosis of a primary headache disorder should be documented along with the diagnosis of MOH.

Medication overuse can exist without MOH: Not every patient who frequently uses an abortive medication develops MOH.

Treatment is multifaceted—and can become complex

Mainstays of treatment of MOH are education about the disorder and detoxification from the overused agent, although specific treatments can differ depending on the agent involved, the frequency and duration of its use, and a patient’s behavioral patterns and psychiatric comorbidities. Often, a daily medication to prevent headache is considered upon, or after, withdrawal of the offending agent. The timing of introducing a preventive might impact its effectiveness. Some refractory cases require more intensive therapy, including hospitalization at a specialized tertiary center.

But before we look at detoxification from an overused agent, it’s important to review one of the best strategies of all in combatting MOH.

Continue to: First and best strategy

Select an appropriate abortive to reduce the risk of MOH. With regard to specific acute headache medications, some nuances other than type of headache should be considered. Nonsteroidal anti-inflammatory drugs (NSAIDs) are recommended as abortive therapy by the American Headache Society for their efficacy, favorable adverse effect profile, and low cost. NSAIDs are protective against development of MOH if a patient’s baseline headache frequency is < 10/mo; at a frequency of 10 to 14 d/mo, however, the risk of MOH increases when using an NSAID.⁶ A similar effect has been seen with triptans.¹⁶ Longer-acting NSAIDs, such as nabumetone and naproxen, have been proposed as less likely to cause MOH, and are even used as bridging therapy sometimes (as long as neither of these was the overused medication).²⁶

The time it takes to develop MOH is shortest with triptans, followed by ergots, then analgesics.²⁷

Prospective cohort studies^6,16 have shown that barbiturates and opioids are more likely to induce MOH; for that reason, agents in these analgesic classes are almost universally avoided unless no other medically acceptable options exist. Using barbiturate-containing compounds or opioids > 4 d/mo exponentially increases the likelihood of MOH.

Promising preclinical data demonstrate that the gepant, or small-molecule calcitonin gene-related peptide (CGRP) receptor antagonist, class of medications used as abortive therapy does not induce medication overuse cutaneous allodynia.²⁸

Provide education. Primary prevention of MOH involves (1) increasing patients’ awareness of how to take medications appropriately and (2) restricting intake of over-the-counter abortive medications. Often, the expert recommendation is to limit abortives to approximately 2 d/wk because more frequent use places patients at risk of further increased use and subsequent MOH.

Continue to: A randomized controlled trial in Norway...

A randomized controlled trial in Norway compared outcomes in 2 groups of patients with MOH: One group was given advice on the disorder by a physician; the other group was not provided with advice. In the “business-as-usual” group, there was no significant improvement; however, when general practitioners provided simple advice (lasting roughly 9 minutes) about reducing abortive medication use to a safe level and cautioned patients that they would be “feeling worse before feeling better,” headache days were reduced by approximately 8 per month and medication days, by 16 per month.²

A subsequent, long-term follow-up study²⁹ of patients from the Norway trial² who had been given advice and education showed a relapse rate (ie, into overuse of headache medication) of only 8% and sustained reduction of headache days and medication use at 16 months.

Offer support and other nondrug interventions. A recent review of 3 studies²³ recommended that extra support for patients from a headache nurse, close follow-up, keeping an electronic diary that provides feedback, and undertaking a short course of psychotherapy can reduce medication overuse and prevent relapse.

If MOH develops, initiate withdrawal, introduce a preventive

Withdraw overused medication. Most current evidence suggests that withdrawal of the offending agent is the most effective factor in reducing headache days and improving quality of life. A randomized controlled trial compared the effects of (1) complete and immediate withdrawal of an abortive medication with (2) reducing its use (ie, limiting intake to 2 d/wk), on headache frequency, disability, and quality of life.³⁰ There was a reduction of headache days in both groups; however, reduction was much greater at 2 months in the complete withdrawal group than in the restricted intake group (respectively, a 41% and a 26% reduction in headache days per month). This effect was sustained at 6 and 12 months in both groups. The study confirmed the results of earlier research^2,15: Abrupt withdrawal leads to reversion to an episodic pattern at 2 to 6 months in approximately 40% to 60% of patients.

More studies are needed to determine the most appropriate treatment course for MOH; however, complete withdrawal of the causative drug is the most important intervention.

Continue to: Consider withdrawal plus preventive treatment

Consider withdrawal plus preventive treatment. Use of sodium valproate, in addition to medication overuse detoxification, led to a significant reduction in headache days and improvement in quality of life at 12 weeks but no difference after 24 weeks, compared with detoxification alone in a randomized, double-blind, placebo-controlled study.³¹

A study of 61 patients showed a larger reduction (by 7.2 d/mo) in headache frequency with any preventive medication in addition to medication withdrawal, compared to withdrawal alone (by 4.1 d/mo) after 3 months; however, the relative benefit was gone at 6 months.³²

A study of 98 patients compared immediate and delayed initiation of preventive medication upon withdrawal of overused abortive medication.³³ Response was defined as a > 50% reduction in headache frequency and was similar in both groups; results showed a 28% response with immediate initiation of a preventive; a 23% response with delayed (ie, 2 months after withdrawal) initiation; and a 48% response in both groups at 12 months.

Collectively, these studies suggest that adding a preventive medication at the time of withdrawal has the potential to reduce headache frequency more quickly than withdrawal alone. However, after 3 to 6 months, the outcome of reduced headache frequency is the same whether or not a preventive medication is used—as long as the offending agent has been withdrawn.

Do preventives work without withdrawing overused medication? Patients with MOH often show little or no improvement with addition of a preventive medication only; their response to a preventive improves after withdrawal of the overused medication. Patients without previous headache improvement after addition of a preventive, who also did not improve 2 months after withdrawal, then demonstrated an overall reduction in headache by 26% when a preventive was reintroduced after withdrawal.²

Continue to: The research evidence for preventives

The research evidence for preventives. Medications for headache prevention have not been extensively evaluated specifically for treating MOH. Here is what’s known:

• Flunarizine, amitriptyline, and beta-blockers usually are ineffective for MOH.²⁴
• Results for topiramate are mixed: A small, double-blind, placebo-controlled chronic migraine study in Europe showed that, in a subgroup of patients with MOH, topiramate led to a small but significant reduction (3.5 d/mo) in headache frequency, compared to placebo.²⁷ A similar study done in the United States did not show a significant difference between the active-treatment and placebo groups.³⁴
• Findings regarding onabotulinumtoxinA are intriguing: In a posthoc analysis of onabotulinumtoxinA to treat chronic migraine, patients with MOH who did not undergo detoxification had an 8 d/mo greater reduction in headache, compared to placebo.³⁵ However, when compared to placebo in conjunction with detoxification, onabotulinumtoxinA demonstrated no benefit.³⁶
• Newer CGRP antagonist and CGRP receptor antagonist monoclonal antibodies are successful preventive medications that have demonstrated a reduction in acute medication use days per month and headache days per month³⁷; these compounds have not been compared to withdrawal alone.

Reducing the severity and duration of withdrawal symptoms

Withdrawal from overused abortive headache medications can lead to worsening headache, nausea, vomiting, hypotension, tachycardia, sleep disturbances, restlessness, anxiety, and nervousness. Symptoms usually last 2 to 10 days but can persist for as long as 4 weeks; duration of withdrawal symptoms varies with the medication that is being overused. In patients who have used a triptan, for example, mean duration of withdrawal is 4.1 days; ergotamine, 6.7 days; and NSAIDs, 9.5 days.²³ Tapered withdrawal is sometimes recommended with opioids and barbiturates to reduce withdrawal symptoms. It is unclear whether starting a preventive medication during withdrawal assists in reducing withdrawal symptoms.³⁸

Bridging therapy to reduce symptoms of withdrawal is often provided despite debatable utility. Available evidence does not favor one agent or method but suggests some strategies that could be helpful:

• A prednisone taper has a potential role during the first 6 days of withdrawal by reducing rebound headache and withdrawal symptoms³⁹; however, oral prednisolone has been shown to have no benefit.⁴⁰
• Alone, IV methylprednisolone seems not to be of benefit; however, in a retrospective study of 94 patients, IV methylprednisolone plus diazepam for 5 days led to a significant reduction in headache frequency and drug consumption that was sustained after 3 months.⁴¹
• Celecoxib was compared to prednisone over a 20-day course: a celecoxib dosage of 400 mg/d for the first 5 days, tapered by 100 mg every 5 days, and an oral prednisone dosage of 75 mg/d for the first 5 days, then tapered every 5 days. Patients taking celecoxib had lower headache intensity but there was no difference in headache frequency and acute medication intake between the groups.⁴²

Other strategies. Using antiemetics and NSAIDs to reduce withdrawal symptoms is widely practiced, but no placebo-­controlled trials have been conducted to support this strategy.

Reduce the risk of medication overuse headache by selecting an appropriate abortive; NSAIDs are recommended for their efficacy, favorable adverse effect profile, and low cost.

Patients in withdrawal might be more likely to benefit from inpatient care if they have a severe comorbidity, such as opioid or barbiturate use; failure to respond to, tolerate, or adhere to treatment; or relapse after withdrawal.³⁸

Continue to: Cognitive behavioral therapy...

Cognitive behavioral therapy, exercise, a headache diary, and biofeedback should be considered in every patient’s treatment strategy because a multidisciplinary approach increases adherence and leads to improvement in headache frequency and a decrease in disability and medication use.⁴³

Predictors of Tx success

A prospective cohort study determined that the rate of MOH relapse is 31% at 6 months, 41% at 1 year, and 45% at 4 years, with the highest risk of relapse during the first year.⁴⁴ Looking at the correlation between type of medication overused and relapse rate, the research indicates that

• triptans have the lowest risk of relapse,⁴⁴
• simple analgesics have a higher risk of relapse than triptans,^22,44 and
• opioids have the highest risk of relapse.²²

Where the data don’t agree. Data on combination analgesics and on ergots are conflicting.²² In addition, data on whether the primary type of headache predicts relapse rate conflict; however, migraine might predict a better outcome than tension-type headache.²²

To recap and expand: Management pearls

The major goals of headache management generally are to rule out secondary headache, reach a correct diagnosis, reduce overall headache frequency, and provide effective abortive medication. A large component of reducing headache frequency is addressing and treating medication overuse.

Seek to understand the nature of the patient’s headache disorder. Components of the history are key in identifying the underlying headache diagnosis and ruling out other, more concerning secondary headache diagnoses. The ICHD-3 is an excellent resource for treating headache disorders because the classification lists specific diagnostic criteria for all recognized headache diagnoses.

Continue to: Medication withdrawal...

Medication withdrawal—with or without preventive medication—should reduce the frequency of MOH in 2 or 3 months. If headache does not become less frequent, however, the headache diagnosis might need to be reconsidered. Minimizing the use of abortive medication is generally recommended, but reduction or withdrawal of these medications does not guarantee that patients will revert to an episodic pattern of headache.

Inpatient care of withdrawal might be beneficial when a patient has a severe comorbidity; does not respond to, tolerate, or adhere to treatment; or relapses after withdrawal.

Treating withdrawal symptoms is a reasonable approach in some patients, but evidence does not support routinely providing bridging therapy.

Apply preventives carefully. Abortive medication withdrawal should generally be completed before initiating preventive medication; however, over the short term, starting preventive therapy while withdrawing the overused medication could assist in reducing headache frequency rapidly. This strategy can put patients at risk of medication adverse effects and using the medications longer than necessary, yet might be reasonable in certain patients, given their comorbidities, risk of relapse, and physician and patient preference. A preventive medication for an individual patient should generally be chosen in line with recommendations of the American Academy of Neurology⁴⁵ and on the basis of the history and comorbidities.

Provide education, which is essential to lowering barriers to success. Patients with MOH must be counseled to understand that (1) a headache treatment that is supposed to be making them feel better is, in fact, making them feel worse and (2) they will get worse before they get better. Many patients are afraid to be without medication to use as needed. It is helpful to educate them on the different types of treatments (abortive, preventive); how MOH interferes with headache prophylaxis and medication efficacy; how MOH alters brain function (ie, aforementioned physiologic changes in pain processing and functional imaging changes²³); and that such change is reversible when medication is withdrawn.

ACKNOWLEDGEMENT
The author thanks Jeffrey Curtis, MD, MPH, for his support and editing assistance with the manuscript.

CORRESPONDENCE
Allison Crain, MD, 2927 N 7th Avenue, Phoenix, AZ 85013; [email protected].

While some guidelines support serotonin norepinephrine reuptake inhibitors (SNRIs) as treatments for back pain, a new systematic review and meta-analysis of existing research found no firm evidence of a benefit. Adverse effects, however, are common.

“Our review shows that, although these medicines are effective, the effect is small and unlikely to be considered clinically important by most patients,” wrote the authors of the review, which appeared Jan. 20 in the BMJ. “Our review also showed that about two-thirds of patients using SNRIs experience adverse events.”

However, the report hinted that certain classes of antidepressants may provide significant relief in knee OA and sciatica.

According to a 2018 review, 10 of 15 clinical guidelines from around the world – including those of the American College of Physicians – recommended antidepressants as treatments for low back pain, and 2 advised against them. “Evidence supporting the use of antidepressants is, however, uncertain,” wrote the authors of the new review, led by Giovanni E. Ferreira, PhD, of the University of Sydney. “Systematic reviews of antidepressants for back pain and osteoarthritis have either not included several published trials, considered only one type of antidepressant (e.g., duloxetine), or failed to assess the certainty of evidence.”

For the new review, the authors analyzed 33 randomized, controlled trials with a total of 5,318 subjects. Both published data and unpublished data from clinical trial registries were included.
 

Back pain trials

A total of 19 trials examined back pain, mostly lower back pain (16 trials), and none lasted more than 1 year. Fifteen examined SNRIs while others looked at other kinds of antidepressants.

The researchers found that “the effect of SNRIs was small [on back pain] and below this review’s predetermined threshold of clinical importance. ... Evidence ranging from low to very low certainty showed no benefit of a range of antidepressant classes, including SSRIs [selective serotonin reuptake inhibitors], tetracyclic antidepressants, SARIs [serotonin antagonist and reuptake inhibitors], and NDRIs [norepinephrine and dopamine reuptake inhibitors] for pain and disability across follow-ups of 2 weeks or less, 3-13 weeks, and 3-12 months.”
 

Sciatica trials

Six trials examined antidepressants as treatments for sciatica. Very-low-certainty evidence suggested that SNRIs reduced pain at up to 2 weeks (1 trial, n = 50) but not at 3-13 weeks (3 trials, n = 96). The results of trials of tricyclic antidepressants (TCAs) were the opposite: low- to very-low-certainty evidence suggested the drugs didn’t reduce pain at up to 2 weeks (2 trials, n = 94) but did at 3-13 weeks (2 trials, n = 114) and 3-12 months (1 trial, n = 60).

“All sciatica trials were small, had imprecise estimates, and were at high risk of bias, which reduced the certainty of evidence to low and very low,” the authors cautioned. “This level of uncertainty indicates that the true estimate of effect of TCAs and SNRIs for sciatica is likely to be substantially different from what we estimated in our review.”
 

Knee OA trials

Eight trials examined SNRIs in knee OA. Moderate-certainty evidence linked the drugs to less pain at up to 2 weeks (four trials, n = 1,328) and low-certainty evidence linked them to less pain at 3-13 weeks (eight trials, n = 1,941). Low-certainty evidence also linked the drugs to less disability at 2 weeks or less (one trial, n = 353) and 3-13 weeks (seven trials, n = 1,810).

In knee OA, “the effect of SNRIs was small and below this review’s predetermined threshold of clinical importance,” the researchers wrote. “However, the lower limit of the confidence interval did contain clinically important effects for pain, but not for disability.”
 

Antidepressant side effects in trials

A total of 21 trials (n = 4,107) looked at side effects when antidepressants were studied as treatments for back pain and OA. Low-certainty evidence in 13 SNRI trials (n = 3,447) suggested a higher risk of any adverse events in antidepressant versus placebo (62.5% vs. 49.7%; relative risk, 1.23, 95% confidence interval, 1.16-1.30), but there was no significantly higher risk of serious adverse events in 10 SNRI trials with 3,309 subjects (1.6% vs. 1.3%; RR, 1.12, 95% CI, 0.61-2.07).

As for adverse effects of non-SNRIs, “the number of studies evaluating the safety of other antidepressant classes was small, trials were underpowered to detect harm, and the certainty of evidence ranged from low to very low,” the researchers wrote.

Going forward, the authors said that “large, definitive randomized trials that are free of industry ties are urgently needed to resolve uncertainties about the efficacy of antidepressants for sciatica and osteoarthritis highlighted by this review.”
 

‘Largely ineffective’ drug treatments

In an accompanying commentary, Martin Underwood, of the University of Warwick in Coventry, England, and Colin Tysall, of the University Hospitals of Coventry and Warwickshire, also in Coventry, noted that “drug treatments are largely ineffective for back pain and osteoarthritis and have the potential for serious harm. We need to work harder to help people with these disorders to live better with their pain without recourse to the prescription pad.”

However, they noted that SNRIs may still be helpful for patients with back pain or OA. “Absolute effect sizes for physical treatments for low-back pain are of similar magnitudes to those reported here and translate into numbers needed to treat of between five and nine. If the same were true for SNRIs, some people might choose to a try that option for a 1 in 10 chance of a worthwhile reduction in pain after 3 months. They can easily stop if treatment is ineffective or does not suit them.”

The research received no specific funding. The review authors disclosed relationships with GlaxoSmithKline (postgraduate scholarship), Pfizer (investigational product for two trials), and Flexeze (provision of heat wraps for a trial). Mr. Underwood reported being a director and shareholder of Clinvivo. Mr. Tysall reported no disclosures.

While some guidelines support serotonin norepinephrine reuptake inhibitors (SNRIs) as treatments for back pain, a new systematic review and meta-analysis of existing research found no firm evidence of a benefit. Adverse effects, however, are common.

“Our review shows that, although these medicines are effective, the effect is small and unlikely to be considered clinically important by most patients,” wrote the authors of the review, which appeared Jan. 20 in the BMJ. “Our review also showed that about two-thirds of patients using SNRIs experience adverse events.”

However, the report hinted that certain classes of antidepressants may provide significant relief in knee OA and sciatica.

According to a 2018 review, 10 of 15 clinical guidelines from around the world – including those of the American College of Physicians – recommended antidepressants as treatments for low back pain, and 2 advised against them. “Evidence supporting the use of antidepressants is, however, uncertain,” wrote the authors of the new review, led by Giovanni E. Ferreira, PhD, of the University of Sydney. “Systematic reviews of antidepressants for back pain and osteoarthritis have either not included several published trials, considered only one type of antidepressant (e.g., duloxetine), or failed to assess the certainty of evidence.”

For the new review, the authors analyzed 33 randomized, controlled trials with a total of 5,318 subjects. Both published data and unpublished data from clinical trial registries were included.
 

Back pain trials

A total of 19 trials examined back pain, mostly lower back pain (16 trials), and none lasted more than 1 year. Fifteen examined SNRIs while others looked at other kinds of antidepressants.

The researchers found that “the effect of SNRIs was small [on back pain] and below this review’s predetermined threshold of clinical importance. ... Evidence ranging from low to very low certainty showed no benefit of a range of antidepressant classes, including SSRIs [selective serotonin reuptake inhibitors], tetracyclic antidepressants, SARIs [serotonin antagonist and reuptake inhibitors], and NDRIs [norepinephrine and dopamine reuptake inhibitors] for pain and disability across follow-ups of 2 weeks or less, 3-13 weeks, and 3-12 months.”
 

Sciatica trials

Six trials examined antidepressants as treatments for sciatica. Very-low-certainty evidence suggested that SNRIs reduced pain at up to 2 weeks (1 trial, n = 50) but not at 3-13 weeks (3 trials, n = 96). The results of trials of tricyclic antidepressants (TCAs) were the opposite: low- to very-low-certainty evidence suggested the drugs didn’t reduce pain at up to 2 weeks (2 trials, n = 94) but did at 3-13 weeks (2 trials, n = 114) and 3-12 months (1 trial, n = 60).

“All sciatica trials were small, had imprecise estimates, and were at high risk of bias, which reduced the certainty of evidence to low and very low,” the authors cautioned. “This level of uncertainty indicates that the true estimate of effect of TCAs and SNRIs for sciatica is likely to be substantially different from what we estimated in our review.”
 

Knee OA trials

Eight trials examined SNRIs in knee OA. Moderate-certainty evidence linked the drugs to less pain at up to 2 weeks (four trials, n = 1,328) and low-certainty evidence linked them to less pain at 3-13 weeks (eight trials, n = 1,941). Low-certainty evidence also linked the drugs to less disability at 2 weeks or less (one trial, n = 353) and 3-13 weeks (seven trials, n = 1,810).

In knee OA, “the effect of SNRIs was small and below this review’s predetermined threshold of clinical importance,” the researchers wrote. “However, the lower limit of the confidence interval did contain clinically important effects for pain, but not for disability.”
 

Antidepressant side effects in trials

A total of 21 trials (n = 4,107) looked at side effects when antidepressants were studied as treatments for back pain and OA. Low-certainty evidence in 13 SNRI trials (n = 3,447) suggested a higher risk of any adverse events in antidepressant versus placebo (62.5% vs. 49.7%; relative risk, 1.23, 95% confidence interval, 1.16-1.30), but there was no significantly higher risk of serious adverse events in 10 SNRI trials with 3,309 subjects (1.6% vs. 1.3%; RR, 1.12, 95% CI, 0.61-2.07).

As for adverse effects of non-SNRIs, “the number of studies evaluating the safety of other antidepressant classes was small, trials were underpowered to detect harm, and the certainty of evidence ranged from low to very low,” the researchers wrote.

Going forward, the authors said that “large, definitive randomized trials that are free of industry ties are urgently needed to resolve uncertainties about the efficacy of antidepressants for sciatica and osteoarthritis highlighted by this review.”
 

‘Largely ineffective’ drug treatments

In an accompanying commentary, Martin Underwood, of the University of Warwick in Coventry, England, and Colin Tysall, of the University Hospitals of Coventry and Warwickshire, also in Coventry, noted that “drug treatments are largely ineffective for back pain and osteoarthritis and have the potential for serious harm. We need to work harder to help people with these disorders to live better with their pain without recourse to the prescription pad.”

However, they noted that SNRIs may still be helpful for patients with back pain or OA. “Absolute effect sizes for physical treatments for low-back pain are of similar magnitudes to those reported here and translate into numbers needed to treat of between five and nine. If the same were true for SNRIs, some people might choose to a try that option for a 1 in 10 chance of a worthwhile reduction in pain after 3 months. They can easily stop if treatment is ineffective or does not suit them.”

The research received no specific funding. The review authors disclosed relationships with GlaxoSmithKline (postgraduate scholarship), Pfizer (investigational product for two trials), and Flexeze (provision of heat wraps for a trial). Mr. Underwood reported being a director and shareholder of Clinvivo. Mr. Tysall reported no disclosures.

While some guidelines support serotonin norepinephrine reuptake inhibitors (SNRIs) as treatments for back pain, a new systematic review and meta-analysis of existing research found no firm evidence of a benefit. Adverse effects, however, are common.

“Our review shows that, although these medicines are effective, the effect is small and unlikely to be considered clinically important by most patients,” wrote the authors of the review, which appeared Jan. 20 in the BMJ. “Our review also showed that about two-thirds of patients using SNRIs experience adverse events.”

However, the report hinted that certain classes of antidepressants may provide significant relief in knee OA and sciatica.

According to a 2018 review, 10 of 15 clinical guidelines from around the world – including those of the American College of Physicians – recommended antidepressants as treatments for low back pain, and 2 advised against them. “Evidence supporting the use of antidepressants is, however, uncertain,” wrote the authors of the new review, led by Giovanni E. Ferreira, PhD, of the University of Sydney. “Systematic reviews of antidepressants for back pain and osteoarthritis have either not included several published trials, considered only one type of antidepressant (e.g., duloxetine), or failed to assess the certainty of evidence.”

For the new review, the authors analyzed 33 randomized, controlled trials with a total of 5,318 subjects. Both published data and unpublished data from clinical trial registries were included.
 

Back pain trials

A total of 19 trials examined back pain, mostly lower back pain (16 trials), and none lasted more than 1 year. Fifteen examined SNRIs while others looked at other kinds of antidepressants.

The researchers found that “the effect of SNRIs was small [on back pain] and below this review’s predetermined threshold of clinical importance. ... Evidence ranging from low to very low certainty showed no benefit of a range of antidepressant classes, including SSRIs [selective serotonin reuptake inhibitors], tetracyclic antidepressants, SARIs [serotonin antagonist and reuptake inhibitors], and NDRIs [norepinephrine and dopamine reuptake inhibitors] for pain and disability across follow-ups of 2 weeks or less, 3-13 weeks, and 3-12 months.”
 

Sciatica trials

Six trials examined antidepressants as treatments for sciatica. Very-low-certainty evidence suggested that SNRIs reduced pain at up to 2 weeks (1 trial, n = 50) but not at 3-13 weeks (3 trials, n = 96). The results of trials of tricyclic antidepressants (TCAs) were the opposite: low- to very-low-certainty evidence suggested the drugs didn’t reduce pain at up to 2 weeks (2 trials, n = 94) but did at 3-13 weeks (2 trials, n = 114) and 3-12 months (1 trial, n = 60).

“All sciatica trials were small, had imprecise estimates, and were at high risk of bias, which reduced the certainty of evidence to low and very low,” the authors cautioned. “This level of uncertainty indicates that the true estimate of effect of TCAs and SNRIs for sciatica is likely to be substantially different from what we estimated in our review.”
 

Knee OA trials

Eight trials examined SNRIs in knee OA. Moderate-certainty evidence linked the drugs to less pain at up to 2 weeks (four trials, n = 1,328) and low-certainty evidence linked them to less pain at 3-13 weeks (eight trials, n = 1,941). Low-certainty evidence also linked the drugs to less disability at 2 weeks or less (one trial, n = 353) and 3-13 weeks (seven trials, n = 1,810).

In knee OA, “the effect of SNRIs was small and below this review’s predetermined threshold of clinical importance,” the researchers wrote. “However, the lower limit of the confidence interval did contain clinically important effects for pain, but not for disability.”
 

Antidepressant side effects in trials

A total of 21 trials (n = 4,107) looked at side effects when antidepressants were studied as treatments for back pain and OA. Low-certainty evidence in 13 SNRI trials (n = 3,447) suggested a higher risk of any adverse events in antidepressant versus placebo (62.5% vs. 49.7%; relative risk, 1.23, 95% confidence interval, 1.16-1.30), but there was no significantly higher risk of serious adverse events in 10 SNRI trials with 3,309 subjects (1.6% vs. 1.3%; RR, 1.12, 95% CI, 0.61-2.07).

As for adverse effects of non-SNRIs, “the number of studies evaluating the safety of other antidepressant classes was small, trials were underpowered to detect harm, and the certainty of evidence ranged from low to very low,” the researchers wrote.

Going forward, the authors said that “large, definitive randomized trials that are free of industry ties are urgently needed to resolve uncertainties about the efficacy of antidepressants for sciatica and osteoarthritis highlighted by this review.”
 

‘Largely ineffective’ drug treatments

In an accompanying commentary, Martin Underwood, of the University of Warwick in Coventry, England, and Colin Tysall, of the University Hospitals of Coventry and Warwickshire, also in Coventry, noted that “drug treatments are largely ineffective for back pain and osteoarthritis and have the potential for serious harm. We need to work harder to help people with these disorders to live better with their pain without recourse to the prescription pad.”

However, they noted that SNRIs may still be helpful for patients with back pain or OA. “Absolute effect sizes for physical treatments for low-back pain are of similar magnitudes to those reported here and translate into numbers needed to treat of between five and nine. If the same were true for SNRIs, some people might choose to a try that option for a 1 in 10 chance of a worthwhile reduction in pain after 3 months. They can easily stop if treatment is ineffective or does not suit them.”

The research received no specific funding. The review authors disclosed relationships with GlaxoSmithKline (postgraduate scholarship), Pfizer (investigational product for two trials), and Flexeze (provision of heat wraps for a trial). Mr. Underwood reported being a director and shareholder of Clinvivo. Mr. Tysall reported no disclosures.

Patients with knee osteoarthritis (OA) who wear stable supportive shoes for 6 months have greater average reductions in knee pain when walking, compared with patients who wear flat flexible shoes, according to a randomized trial that included more than 160 patients.

copyright Nandyphotos/Thinkstock

“Contrary to our hypothesis, flat flexible shoes were not superior to stable supportive shoes,” reported Kade L. Paterson, PhD, of the University of Melbourne, and colleagues. Their study was published Jan. 12 in Annals of Internal Medicine.
 

Research gap

Abnormal knee joint loading has been implicated in the pathogenesis of knee OA. Guidelines recommend that patients wear appropriate footwear, but research has not established which shoes are best.

The 2019 American College of Rheumatology clinical guidelines note that “optimal footwear is likely to be of considerable importance for those with knee and/or hip OA,” but “the available studies do not define the best type of footwear to improve specific outcomes for knee or hip OA.”

Some doctors call for thick, shock-absorbing soles and arch supports, based on expert opinion. On the other hand, studies have found that knee loading is lower with flat flexible shoes, and preliminary evidence has suggested that flat flexible shoes may improve OA symptoms, the investigators said.

To study this question, they enrolled in their trial 164 patients aged 50 years and older who had radiographic medial knee OA. Participants had knee pain on most days of the previous month, tibiofemoral osteophytes, and moderate to severe tibiofemoral OA.

The researchers randomly assigned 82 participants to flat flexible shoes and 82 participants to stable supportive shoes, worn for at least 6 hours a day for 6 months.

In the trial, flat flexible shoes included Merrell Bare Access (men’s and women’s), Vivobarefoot Primus Lite (men’s and women’s), Vivobarefoot Mata Canvas (men’s), Converse Dainty Low (women’s), and Lacoste Marice (men’s).

Stable supportive shoes included ASICS Kayano (men’s and women’s), Merrell Jungle Moc (men’s), Nike Air Max 90 Ultra (women’s), Rockport Edge Hill (men’s), and New Balance 624 (women’s).

After participants were randomly assigned to a group, they chose two different pairs of shoes from their assigned footwear group.

“Participants were not told that the purpose of the study was to compare flat flexible with stable supportive shoes,” the researchers noted. “Instead, they were informed only that the trial was comparing the effects of ‘different shoes’ on knee OA symptoms.”

The primary outcomes were changes in walking pain on a 0-10 scale and physical function as assessed by the Western Ontario and McMaster Universities Osteoarthritis Index subscale at 6 months. The researchers also assessed other measures of pain and function, physical activity, and quality of life.

In all, 161 participants reported 6-month primary outcomes. The between-group difference in change in pain favored stable supportive shoes (mean difference, 1.1 units). In the flat flexible shoe group, overall average knee pain while walking decreased from 6.3 at baseline to 5.2 at 6 months. In the stable supportive shoe group, knee pain while walking decreased from 6.1 to 4.

In addition, improvements in knee-related quality of life and ipsilateral hip pain favored stable supportive shoes.

Participants who wore stable supportive shoes also were less likely to report adverse events, compared with those who wore flat flexible shoes (15% vs. 32%). Knee pain, ankle or foot pain, and shin or calf pain were among the adverse events reported.
 

‘Important work’

“This study suggests that more supportive shoes may help some patients with knee osteoarthritis feel better,” Constance R. Chu, MD, professor of orthopedic surgery at Stanford (Calif.) University, said in an interview. “Shoes, insoles, wedges, and high heels have been shown to change loading of the knee related to knee pain and osteoarthritis ... This is important work toward providing more specific information on the optimum shoes for people with different patterns and types of arthritis to reduce pain and disability from early knee OA.”

The reported changes in pain may be clinically meaningful for many but not all patients, the authors wrote. “Despite biomechanical evidence showing that flat flexible shoes reduce medial knee load compared with stable supportive shoes, our findings show that this does not translate to improved knee osteoarthritis symptoms,” they said. “This may be because relationships between knee loading and symptoms are not as strong as previously thought, or because the small reductions in medial knee load with flat flexible shoes are insufficient to substantively improve pain and function.”

The trial did not include a control group of patients who wore their usual shoes, and it focused on a select subgroup of patients with knee OA, which may limit the study’s generalizability, the authors noted. The study excluded people with lateral joint space narrowing greater than or equal to medial, those with recent or planned knee surgery, and those who were using shoe orthoses or customized shoes.

The study was supported by grants from the National Health and Medical Research Council. Dr. Chu had no relevant disclosures.

[email protected]

Patients with knee osteoarthritis (OA) who wear stable supportive shoes for 6 months have greater average reductions in knee pain when walking, compared with patients who wear flat flexible shoes, according to a randomized trial that included more than 160 patients.

copyright Nandyphotos/Thinkstock

“Contrary to our hypothesis, flat flexible shoes were not superior to stable supportive shoes,” reported Kade L. Paterson, PhD, of the University of Melbourne, and colleagues. Their study was published Jan. 12 in Annals of Internal Medicine.
 

Research gap

Abnormal knee joint loading has been implicated in the pathogenesis of knee OA. Guidelines recommend that patients wear appropriate footwear, but research has not established which shoes are best.

The 2019 American College of Rheumatology clinical guidelines note that “optimal footwear is likely to be of considerable importance for those with knee and/or hip OA,” but “the available studies do not define the best type of footwear to improve specific outcomes for knee or hip OA.”

Some doctors call for thick, shock-absorbing soles and arch supports, based on expert opinion. On the other hand, studies have found that knee loading is lower with flat flexible shoes, and preliminary evidence has suggested that flat flexible shoes may improve OA symptoms, the investigators said.

To study this question, they enrolled in their trial 164 patients aged 50 years and older who had radiographic medial knee OA. Participants had knee pain on most days of the previous month, tibiofemoral osteophytes, and moderate to severe tibiofemoral OA.

The researchers randomly assigned 82 participants to flat flexible shoes and 82 participants to stable supportive shoes, worn for at least 6 hours a day for 6 months.

In the trial, flat flexible shoes included Merrell Bare Access (men’s and women’s), Vivobarefoot Primus Lite (men’s and women’s), Vivobarefoot Mata Canvas (men’s), Converse Dainty Low (women’s), and Lacoste Marice (men’s).

Stable supportive shoes included ASICS Kayano (men’s and women’s), Merrell Jungle Moc (men’s), Nike Air Max 90 Ultra (women’s), Rockport Edge Hill (men’s), and New Balance 624 (women’s).

After participants were randomly assigned to a group, they chose two different pairs of shoes from their assigned footwear group.

“Participants were not told that the purpose of the study was to compare flat flexible with stable supportive shoes,” the researchers noted. “Instead, they were informed only that the trial was comparing the effects of ‘different shoes’ on knee OA symptoms.”

The primary outcomes were changes in walking pain on a 0-10 scale and physical function as assessed by the Western Ontario and McMaster Universities Osteoarthritis Index subscale at 6 months. The researchers also assessed other measures of pain and function, physical activity, and quality of life.

In all, 161 participants reported 6-month primary outcomes. The between-group difference in change in pain favored stable supportive shoes (mean difference, 1.1 units). In the flat flexible shoe group, overall average knee pain while walking decreased from 6.3 at baseline to 5.2 at 6 months. In the stable supportive shoe group, knee pain while walking decreased from 6.1 to 4.

In addition, improvements in knee-related quality of life and ipsilateral hip pain favored stable supportive shoes.

Participants who wore stable supportive shoes also were less likely to report adverse events, compared with those who wore flat flexible shoes (15% vs. 32%). Knee pain, ankle or foot pain, and shin or calf pain were among the adverse events reported.
 

‘Important work’

“This study suggests that more supportive shoes may help some patients with knee osteoarthritis feel better,” Constance R. Chu, MD, professor of orthopedic surgery at Stanford (Calif.) University, said in an interview. “Shoes, insoles, wedges, and high heels have been shown to change loading of the knee related to knee pain and osteoarthritis ... This is important work toward providing more specific information on the optimum shoes for people with different patterns and types of arthritis to reduce pain and disability from early knee OA.”

The reported changes in pain may be clinically meaningful for many but not all patients, the authors wrote. “Despite biomechanical evidence showing that flat flexible shoes reduce medial knee load compared with stable supportive shoes, our findings show that this does not translate to improved knee osteoarthritis symptoms,” they said. “This may be because relationships between knee loading and symptoms are not as strong as previously thought, or because the small reductions in medial knee load with flat flexible shoes are insufficient to substantively improve pain and function.”

The trial did not include a control group of patients who wore their usual shoes, and it focused on a select subgroup of patients with knee OA, which may limit the study’s generalizability, the authors noted. The study excluded people with lateral joint space narrowing greater than or equal to medial, those with recent or planned knee surgery, and those who were using shoe orthoses or customized shoes.

The study was supported by grants from the National Health and Medical Research Council. Dr. Chu had no relevant disclosures.

[email protected]

Patients with knee osteoarthritis (OA) who wear stable supportive shoes for 6 months have greater average reductions in knee pain when walking, compared with patients who wear flat flexible shoes, according to a randomized trial that included more than 160 patients.

copyright Nandyphotos/Thinkstock

“Contrary to our hypothesis, flat flexible shoes were not superior to stable supportive shoes,” reported Kade L. Paterson, PhD, of the University of Melbourne, and colleagues. Their study was published Jan. 12 in Annals of Internal Medicine.
 

Research gap

Abnormal knee joint loading has been implicated in the pathogenesis of knee OA. Guidelines recommend that patients wear appropriate footwear, but research has not established which shoes are best.

The 2019 American College of Rheumatology clinical guidelines note that “optimal footwear is likely to be of considerable importance for those with knee and/or hip OA,” but “the available studies do not define the best type of footwear to improve specific outcomes for knee or hip OA.”

Some doctors call for thick, shock-absorbing soles and arch supports, based on expert opinion. On the other hand, studies have found that knee loading is lower with flat flexible shoes, and preliminary evidence has suggested that flat flexible shoes may improve OA symptoms, the investigators said.

To study this question, they enrolled in their trial 164 patients aged 50 years and older who had radiographic medial knee OA. Participants had knee pain on most days of the previous month, tibiofemoral osteophytes, and moderate to severe tibiofemoral OA.

The researchers randomly assigned 82 participants to flat flexible shoes and 82 participants to stable supportive shoes, worn for at least 6 hours a day for 6 months.

In the trial, flat flexible shoes included Merrell Bare Access (men’s and women’s), Vivobarefoot Primus Lite (men’s and women’s), Vivobarefoot Mata Canvas (men’s), Converse Dainty Low (women’s), and Lacoste Marice (men’s).

Stable supportive shoes included ASICS Kayano (men’s and women’s), Merrell Jungle Moc (men’s), Nike Air Max 90 Ultra (women’s), Rockport Edge Hill (men’s), and New Balance 624 (women’s).

After participants were randomly assigned to a group, they chose two different pairs of shoes from their assigned footwear group.

“Participants were not told that the purpose of the study was to compare flat flexible with stable supportive shoes,” the researchers noted. “Instead, they were informed only that the trial was comparing the effects of ‘different shoes’ on knee OA symptoms.”

The primary outcomes were changes in walking pain on a 0-10 scale and physical function as assessed by the Western Ontario and McMaster Universities Osteoarthritis Index subscale at 6 months. The researchers also assessed other measures of pain and function, physical activity, and quality of life.

In all, 161 participants reported 6-month primary outcomes. The between-group difference in change in pain favored stable supportive shoes (mean difference, 1.1 units). In the flat flexible shoe group, overall average knee pain while walking decreased from 6.3 at baseline to 5.2 at 6 months. In the stable supportive shoe group, knee pain while walking decreased from 6.1 to 4.

In addition, improvements in knee-related quality of life and ipsilateral hip pain favored stable supportive shoes.

Participants who wore stable supportive shoes also were less likely to report adverse events, compared with those who wore flat flexible shoes (15% vs. 32%). Knee pain, ankle or foot pain, and shin or calf pain were among the adverse events reported.
 

‘Important work’

“This study suggests that more supportive shoes may help some patients with knee osteoarthritis feel better,” Constance R. Chu, MD, professor of orthopedic surgery at Stanford (Calif.) University, said in an interview. “Shoes, insoles, wedges, and high heels have been shown to change loading of the knee related to knee pain and osteoarthritis ... This is important work toward providing more specific information on the optimum shoes for people with different patterns and types of arthritis to reduce pain and disability from early knee OA.”

The reported changes in pain may be clinically meaningful for many but not all patients, the authors wrote. “Despite biomechanical evidence showing that flat flexible shoes reduce medial knee load compared with stable supportive shoes, our findings show that this does not translate to improved knee osteoarthritis symptoms,” they said. “This may be because relationships between knee loading and symptoms are not as strong as previously thought, or because the small reductions in medial knee load with flat flexible shoes are insufficient to substantively improve pain and function.”

The trial did not include a control group of patients who wore their usual shoes, and it focused on a select subgroup of patients with knee OA, which may limit the study’s generalizability, the authors noted. The study excluded people with lateral joint space narrowing greater than or equal to medial, those with recent or planned knee surgery, and those who were using shoe orthoses or customized shoes.

The study was supported by grants from the National Health and Medical Research Council. Dr. Chu had no relevant disclosures.

[email protected]

Hospitalized patients with Ehlers-Danlos syndrome (EDS) are more likely to have gastrointestinal, cardiovascular, autonomic, and allergic disorders than are hospitalized patients who do not have EDS, according to a new study of hospital outcomes in these four areas.

“Further research is necessary to explore the prevalence of these manifestations in the different subtypes of EDS and in outpatient population,” wrote Rachel S. Brooks of the University of Connecticut, Farmington, and her coauthors. The study was published in Rheumatology.

To investigate previously observed connections between EDS and these four types of complications, the researchers launched a case-control study using hospital records from the 2016 National Inpatient Sample. A total of 2,007 patients with EDS were identified via ICD-10 code and matched with 4,014 non-EDS patients according to 5-year age intervals, sex, and month of admission. EDS patients had an average age of nearly 37, and 84% were female. The average hospitalization was lengthier for EDS patients (4.77 days) than for controls (4.07 days).

GI conditions were found in 44% of EDS patients, compared with 18% of controls (odds ratio, 3.57; 95% confidence interval, 3.17-4.02; P < .0001). Among the more likely conditions were functional disorders of the stomach (OR, 5.18; 95% CI, 2.16-12.42; P < .0001), unspecified abdominal pain (OR, 3.97; 95% CI, 2.34-6.73; P < .0001), irritable bowel syndrome (OR, 7.44; 95% CI, 5.07-10.94; P < .0001), and nausea (OR, 3.20; 95% CI, 1.95-5.24; P < .0001).

Autonomic dysfunction was found in 20% of EDS patients, compared with 6% of controls (OR, 4.45; 95% CI, 3.71-5.32; P < .0001). They were significantly more likely to have postural orthostatic tachycardia syndrome (OR, 223.77; 95% CI, 31.21-1604.46; P < .0001), orthostatic hypotension (OR, 8.98; 95% CI, 5.36-15.03; P < .0001), syncope (OR, 3.62; 95% CI, 2.23-5.82; P < .0001), and other autonomic nervous system disorders (OR, 54.72; 95% CI, 7.43-403.00; P < .0001).

Food allergies were also considerably more likely to occur in EDS patients (OR, 3.88; 95% CI, 2.65-5.66; P < .0001), as were cardiovascular complications like mitral valve disorders, aortic aneurysm, and cardiac dysrhythmias (OR, 6.16; 95% CI, 4.60-8.23; P < .0001). Although EDS patients were more likely to have hospital stays that lasted longer than 4 days, there was no notable difference in mortality (OR, 0.79; 95% CI, 0.41-1.50; P = .47).

After multivariate regression analysis that adjusted for age, sex, race, and smoking status, EDS patients were more likely to have GI (OR, 3.53; 95% CI, 3.08-4.03; P < .0001), autonomic (OR, 4.13; 95% CI, 3.40-5.01; P < .0001), allergic (OR, 3.92; 95% CI, 2.57-5.98; P < .0001), and cardiovascular complications (OR, 5.82; 95% CI, 4.21-8.03; P < .0001).
 

Shining a much-needed light on the conditions associated with EDS

“Anyone who takes care of patients with EDS has likely seen some of these complications before and knows they can occur,” Jordan T. Jones, DO, a pediatric rheumatologist at Children’s Mercy Hospital in Kansas City, Mo., said in an interview. “I think this study legitimizes what many who take care of patients with EDS know to be true, and for those who don’t, it brings a lot of attention to many of the symptoms and associated conditions.”

He did, however, draw a conclusion that differed from one of the researchers’ chief observations.

“They note that these patients have a longer-than-average hospital stay, suggesting that EDS may be linked to adverse complications during hospitalization,” he said. “I think the reason for longer-than-average hospital stays is due to the number of symptoms and complexity of these patients, which can lead to delays in diagnosis. The complexity can lead to more involved evaluation that keeps them in the hospital longer than usual. Another reason for longer-than-average hospital stays that I’ve seen is the presentation of severe and chronic pain, which can be difficult to treat in the hospital and then transition to outpatient therapy. An inpatient hospitalization is not always the best place to treat chronic pain symptoms, which can drag out a hospital stay.”

He also highlighted the lack of discussion regarding musculoskeletal complications, which he sees as one of the most common symptoms related to EDS.

“As a rheumatologist, I see many patients with EDS present with chronic pain, chronic muscle weakness, and chronic fatigue. If you think about the joint laxity with EDS, these patients are a perfect setup to develop tight, weak muscles, which leads to a lot of musculoskeletal pain and fatigue.”

That said, he ultimately emphasized the clear benefits of such a large study on such an under-researched subject.

“We think EDS is more common than is reported,” he said. “But despite that, there are still a lot of people who don’t know about EDS, understand it, or appreciate how to evaluate for it. One of the best things this study does is bring more visibility to this disease and the associated conditions related to it.”

The authors declared no potential conflicts of interest.

SOURCE: Brooks RS et al. Rheumatology. 2021 Jan 7. doi: 10.1093/rheumatology/keaa926.

Hospitalized patients with Ehlers-Danlos syndrome (EDS) are more likely to have gastrointestinal, cardiovascular, autonomic, and allergic disorders than are hospitalized patients who do not have EDS, according to a new study of hospital outcomes in these four areas.

“Further research is necessary to explore the prevalence of these manifestations in the different subtypes of EDS and in outpatient population,” wrote Rachel S. Brooks of the University of Connecticut, Farmington, and her coauthors. The study was published in Rheumatology.

To investigate previously observed connections between EDS and these four types of complications, the researchers launched a case-control study using hospital records from the 2016 National Inpatient Sample. A total of 2,007 patients with EDS were identified via ICD-10 code and matched with 4,014 non-EDS patients according to 5-year age intervals, sex, and month of admission. EDS patients had an average age of nearly 37, and 84% were female. The average hospitalization was lengthier for EDS patients (4.77 days) than for controls (4.07 days).

GI conditions were found in 44% of EDS patients, compared with 18% of controls (odds ratio, 3.57; 95% confidence interval, 3.17-4.02; P < .0001). Among the more likely conditions were functional disorders of the stomach (OR, 5.18; 95% CI, 2.16-12.42; P < .0001), unspecified abdominal pain (OR, 3.97; 95% CI, 2.34-6.73; P < .0001), irritable bowel syndrome (OR, 7.44; 95% CI, 5.07-10.94; P < .0001), and nausea (OR, 3.20; 95% CI, 1.95-5.24; P < .0001).

Autonomic dysfunction was found in 20% of EDS patients, compared with 6% of controls (OR, 4.45; 95% CI, 3.71-5.32; P < .0001). They were significantly more likely to have postural orthostatic tachycardia syndrome (OR, 223.77; 95% CI, 31.21-1604.46; P < .0001), orthostatic hypotension (OR, 8.98; 95% CI, 5.36-15.03; P < .0001), syncope (OR, 3.62; 95% CI, 2.23-5.82; P < .0001), and other autonomic nervous system disorders (OR, 54.72; 95% CI, 7.43-403.00; P < .0001).

Food allergies were also considerably more likely to occur in EDS patients (OR, 3.88; 95% CI, 2.65-5.66; P < .0001), as were cardiovascular complications like mitral valve disorders, aortic aneurysm, and cardiac dysrhythmias (OR, 6.16; 95% CI, 4.60-8.23; P < .0001). Although EDS patients were more likely to have hospital stays that lasted longer than 4 days, there was no notable difference in mortality (OR, 0.79; 95% CI, 0.41-1.50; P = .47).

After multivariate regression analysis that adjusted for age, sex, race, and smoking status, EDS patients were more likely to have GI (OR, 3.53; 95% CI, 3.08-4.03; P < .0001), autonomic (OR, 4.13; 95% CI, 3.40-5.01; P < .0001), allergic (OR, 3.92; 95% CI, 2.57-5.98; P < .0001), and cardiovascular complications (OR, 5.82; 95% CI, 4.21-8.03; P < .0001).
 

Shining a much-needed light on the conditions associated with EDS

“Anyone who takes care of patients with EDS has likely seen some of these complications before and knows they can occur,” Jordan T. Jones, DO, a pediatric rheumatologist at Children’s Mercy Hospital in Kansas City, Mo., said in an interview. “I think this study legitimizes what many who take care of patients with EDS know to be true, and for those who don’t, it brings a lot of attention to many of the symptoms and associated conditions.”

He did, however, draw a conclusion that differed from one of the researchers’ chief observations.

“They note that these patients have a longer-than-average hospital stay, suggesting that EDS may be linked to adverse complications during hospitalization,” he said. “I think the reason for longer-than-average hospital stays is due to the number of symptoms and complexity of these patients, which can lead to delays in diagnosis. The complexity can lead to more involved evaluation that keeps them in the hospital longer than usual. Another reason for longer-than-average hospital stays that I’ve seen is the presentation of severe and chronic pain, which can be difficult to treat in the hospital and then transition to outpatient therapy. An inpatient hospitalization is not always the best place to treat chronic pain symptoms, which can drag out a hospital stay.”

He also highlighted the lack of discussion regarding musculoskeletal complications, which he sees as one of the most common symptoms related to EDS.

“As a rheumatologist, I see many patients with EDS present with chronic pain, chronic muscle weakness, and chronic fatigue. If you think about the joint laxity with EDS, these patients are a perfect setup to develop tight, weak muscles, which leads to a lot of musculoskeletal pain and fatigue.”

That said, he ultimately emphasized the clear benefits of such a large study on such an under-researched subject.

“We think EDS is more common than is reported,” he said. “But despite that, there are still a lot of people who don’t know about EDS, understand it, or appreciate how to evaluate for it. One of the best things this study does is bring more visibility to this disease and the associated conditions related to it.”

The authors declared no potential conflicts of interest.

SOURCE: Brooks RS et al. Rheumatology. 2021 Jan 7. doi: 10.1093/rheumatology/keaa926.

Hospitalized patients with Ehlers-Danlos syndrome (EDS) are more likely to have gastrointestinal, cardiovascular, autonomic, and allergic disorders than are hospitalized patients who do not have EDS, according to a new study of hospital outcomes in these four areas.

“Further research is necessary to explore the prevalence of these manifestations in the different subtypes of EDS and in outpatient population,” wrote Rachel S. Brooks of the University of Connecticut, Farmington, and her coauthors. The study was published in Rheumatology.

To investigate previously observed connections between EDS and these four types of complications, the researchers launched a case-control study using hospital records from the 2016 National Inpatient Sample. A total of 2,007 patients with EDS were identified via ICD-10 code and matched with 4,014 non-EDS patients according to 5-year age intervals, sex, and month of admission. EDS patients had an average age of nearly 37, and 84% were female. The average hospitalization was lengthier for EDS patients (4.77 days) than for controls (4.07 days).

GI conditions were found in 44% of EDS patients, compared with 18% of controls (odds ratio, 3.57; 95% confidence interval, 3.17-4.02; P < .0001). Among the more likely conditions were functional disorders of the stomach (OR, 5.18; 95% CI, 2.16-12.42; P < .0001), unspecified abdominal pain (OR, 3.97; 95% CI, 2.34-6.73; P < .0001), irritable bowel syndrome (OR, 7.44; 95% CI, 5.07-10.94; P < .0001), and nausea (OR, 3.20; 95% CI, 1.95-5.24; P < .0001).

Autonomic dysfunction was found in 20% of EDS patients, compared with 6% of controls (OR, 4.45; 95% CI, 3.71-5.32; P < .0001). They were significantly more likely to have postural orthostatic tachycardia syndrome (OR, 223.77; 95% CI, 31.21-1604.46; P < .0001), orthostatic hypotension (OR, 8.98; 95% CI, 5.36-15.03; P < .0001), syncope (OR, 3.62; 95% CI, 2.23-5.82; P < .0001), and other autonomic nervous system disorders (OR, 54.72; 95% CI, 7.43-403.00; P < .0001).

Food allergies were also considerably more likely to occur in EDS patients (OR, 3.88; 95% CI, 2.65-5.66; P < .0001), as were cardiovascular complications like mitral valve disorders, aortic aneurysm, and cardiac dysrhythmias (OR, 6.16; 95% CI, 4.60-8.23; P < .0001). Although EDS patients were more likely to have hospital stays that lasted longer than 4 days, there was no notable difference in mortality (OR, 0.79; 95% CI, 0.41-1.50; P = .47).

After multivariate regression analysis that adjusted for age, sex, race, and smoking status, EDS patients were more likely to have GI (OR, 3.53; 95% CI, 3.08-4.03; P < .0001), autonomic (OR, 4.13; 95% CI, 3.40-5.01; P < .0001), allergic (OR, 3.92; 95% CI, 2.57-5.98; P < .0001), and cardiovascular complications (OR, 5.82; 95% CI, 4.21-8.03; P < .0001).
 

Shining a much-needed light on the conditions associated with EDS

“Anyone who takes care of patients with EDS has likely seen some of these complications before and knows they can occur,” Jordan T. Jones, DO, a pediatric rheumatologist at Children’s Mercy Hospital in Kansas City, Mo., said in an interview. “I think this study legitimizes what many who take care of patients with EDS know to be true, and for those who don’t, it brings a lot of attention to many of the symptoms and associated conditions.”

He did, however, draw a conclusion that differed from one of the researchers’ chief observations.

“They note that these patients have a longer-than-average hospital stay, suggesting that EDS may be linked to adverse complications during hospitalization,” he said. “I think the reason for longer-than-average hospital stays is due to the number of symptoms and complexity of these patients, which can lead to delays in diagnosis. The complexity can lead to more involved evaluation that keeps them in the hospital longer than usual. Another reason for longer-than-average hospital stays that I’ve seen is the presentation of severe and chronic pain, which can be difficult to treat in the hospital and then transition to outpatient therapy. An inpatient hospitalization is not always the best place to treat chronic pain symptoms, which can drag out a hospital stay.”

He also highlighted the lack of discussion regarding musculoskeletal complications, which he sees as one of the most common symptoms related to EDS.

“As a rheumatologist, I see many patients with EDS present with chronic pain, chronic muscle weakness, and chronic fatigue. If you think about the joint laxity with EDS, these patients are a perfect setup to develop tight, weak muscles, which leads to a lot of musculoskeletal pain and fatigue.”

That said, he ultimately emphasized the clear benefits of such a large study on such an under-researched subject.

“We think EDS is more common than is reported,” he said. “But despite that, there are still a lot of people who don’t know about EDS, understand it, or appreciate how to evaluate for it. One of the best things this study does is bring more visibility to this disease and the associated conditions related to it.”

The authors declared no potential conflicts of interest.

SOURCE: Brooks RS et al. Rheumatology. 2021 Jan 7. doi: 10.1093/rheumatology/keaa926.

In the midst of an opioid overdose public health crisis, the US Department of Health and Human Services developed a 5-point strategy to combat this problem. One aspect of this strategy is improved pain management.¹ There is high demand for pain management services with a limited number of health care professionals appropriately trained to deliver care.² Pharmacists are integral members of the interdisciplinary pain team and meet this demand.

Background

For almost 50 years, pharmacists at the US Department of Veterans Affairs (VA) have been functioning as advanced practice providers (APP).³ Clinical pharmacy specialists (CPS) provide comprehensive medication management (CMM) and have a scope of practice (SOP). The SOP serves as the collaborating agreement and outlines the clinical duties permitted in delivering patient care. In addition, the SOP may indicate specific practice areas and are standardized across VA (Table 1).^4,5 Pharmacists apply for a SOP and must prove their competency in the practice area and provide documentation of their education, training, experience, knowledge, and skills.^5,6 Residency and/or board certification are not required though helpful. A pharmacist’s SOP is reviewed and approved by the facility executive committee.⁵ Pharmacists with a SOP undergo professional practice evaluation twice a year. Prescribing controlled substances is permissible in the SOP if approved by the facility and allowed by the state of licensure. According to the US Drug Enforcement Agency (DEA) as of February 10, 2020, 8 states (California, Washington, Idaho, Massachusetts, Montana, New Mexico, North Carolina, and Ohio) allow pharmacists to prescribe controlled substances.⁷

The VA developed the Pharmacists Achieve Results with Medications Documentation (PhARMD) tool that allows clinical pharmacists to document specific interventions made during clinical care and is included in their progress note. Data from fiscal year 2017 demonstrates that 136,041 pain management interventions were made by pharmacists across VA. The majority of these interventions were implemented by a CPS working autonomously as an APP.⁸

Several articles discuss the pharmacists role in the opioid crisis, although no outcomes data were provided. Chisholm-Burns and colleagues listed multiple potential ways that pharmacists can intervene, including managing pain in primary care clinic settings by using collaborative drug therapy agreements (CDTAs), using opioid exit plans and discharge planning in collaboration with other health care providers (HCPs), or making recommendations to the prescribers before writing prescriptions.⁹ Compton and colleagues similarly reviewed pharmacist roles in the opioid crisis. However, their focus was on dispensing pharmacists that provided education to patients about storage and disposal of opioids, identified opioid misuse, provided opioid overdose education and naloxone, and checked prescription drug monitoring programs (PDMPs).¹⁰ Missing from these articles was the role of the clinical pharmacist working as an APP delivering direct patient care and prescribing controlled substances.

Hammer and colleagues discussed the role of an oncology CPS with controlled substance prescriptive authority in pain management at an outpatient cancer center in Washington state.¹¹ Under a CDTA, pharmacists could prescribe medications, including controlled substances if they obtain DEA registration. The pharmacist completed a comprehensive in-person assessment. The attending physician conducted a physical examination. Then the pharmacist presented the patient and proposed regimen to the interprofessional team to determine a final plan. Ultimately, the pharmacist wrote any controlled substance prescriptions. The patient followed up every 1 to 4 weeks by telephone with a nurse, and in-person assessments occurred at least every 6 months. No outcomes data were provided.¹¹

Dole and colleagues reviewed the role of a pharmacist who had controlled substance prescriptive authority in a pain management clinic. The pharmacist provider saw up to 18 patients a day and then managed refill requests for 3 hours a day. The main outcome was change in visual analog scale (VAS) pain scores. Findings showed that reductions in VAS pain scores were statistically significant (P < .01). The pharmacist processed about 150 refills with an unclear number of controlled substances requests a day based on a medication-refill protocol. This was felt to improve access to physicians for acute needs, improve consistency in refills, and capture patients in need of follow-up. Additionally, the clinic saved $455,238 after 1 year.¹²

Study Aims

A review of the literature indicated sparse data on the impact of a pharmacist on opioid tapering, opioid dose, and opioid risk mitigation when the pharmacist is prescribing controlled substances. The purpose of this retrospective review was to characterize the controlled substance prescribing practices by the pharmacy pain clinic. The aim was to examine the pharmacist impact on morphine milligram equivalent (MME) and compliance with opioid risk mitigation strategies.

Methods

This project was a retrospective, single-center, chart review. The project was reviewed and approved by the University of Missouri-Columbia Institutional Review Board used by the Harry S. Truman Memorial Veterans’ Hospital (HSTMVH) as a quality improvement project. The author applied for controlled substance registration through the DEA and was issued registration April 30, 2018. The State of Ohio Board of Pharmacy was contacted as required by Ohio Administrative Code. The author's updated SOP to allow controlled substance prescribing was approved July 23, 2018. The CPS functions as an APP within an interdisciplinary pain management team that includes physicians, occupational and physical therapists, complementary and integrative health, and a psychologist. The reason for Pharmacy Pain Consult is required and it is primarily submitted through the electronic health record. The consult is reviewed for appropriateness and once approved is scheduled by support staff. Once the patient is stabilized, the patient is discharged back to their primary care provider (PCP) or referring provider for continued care. Patients were considered stabilized when their patient-specific goals were met, which varied from use of the lowest effective opioid dose to taper to discontinuation of opioids with no further medication changes needed. The taper strategy for each patient was individualized. Patients were generally tapered on their existing opioid medication unless they were new to the VA and on nonformulary medications or experiencing a significant adverse reaction. Numerous references are available through VA to assist with opioid tapering.^13,14 The CPS is able to refer patients to other services, including behavioral health for substance use disorder treatment and medication-assisted treatment if concerns were identified.

Initial data were collected from the Veterans Integrated Service Network (VISN) 15 Corporate Data Warehouse by the VISN pharmacy analytics program manager. The original report included patients prescribed a Schedule II to V controlled substance by the author from July 1, 2018 to January 31, 2020. Chart review was conducted on each patient to obtain additional data. At the time of consult and discharge the following data were collected: opioid medication; MME; use of opioid risk mitigation strategies, such as urine drug screens (UDS), informed consent, opioid overdose education and naloxone distribution program (OEND), risk assessment via stratification tool for opioid risk mitigation (STORM), PDMP checks; and nonopioid medication number and classes.

Patients were included in the review if they were prescribed an opioid Schedule II or III controlled substance between July 1, 2018 and January 31, 2020. Patient were excluded if they were prescribed an opioid Schedule II or III controlled substance primarily as coverage for another prescriber. Patients prescribed only pregabalin, tramadol, or a benzodiazepine also were excluded.

The primary endpoint was change in MME from baseline to discharge from clinic. Secondary endpoints included change in opioid risk mitigation strategies and change in opioid medications prescribed from baseline to discharge.

Descriptive statistics were used to analyze parts of the data. A 2-sided t test was used to compare baseline and discharge MME. The Fisher exact test was used to compare nominal data of opioid risk mitigation strategies.

Calculation of MME was performed using the conversion factors provided by the Centers Disease Control and Prevention (CDC) for opioid guideline.¹⁵ For buprenorphine, tapentadol, and levorphanol conversion ratios were obtained from other sources. The conversion ratios used, included 75:1 for oral morphine to transdermal buprenorphine, 1:3.3 for oral morphine to oral tapentadol, and 1:7.5 for oral levorphanol to oral morphine.^16,17 The Revised Standards for Quality Improvement Reporting Excellence (SQUIRE 2.0) was used to write the manuscript.¹⁸

Results

Seventy-five patients were included in this review. The average age of patients was 66 years; and 12% were female (n = 9) (Table 2). The largest number of consults came from PCPs (44%, n = 33) and the pain clinic (43%, n = 32). Nearly half (48%) of the consultations were for opioid tapering (n = 36), followed by 37% for opioid optimization or monitoring (n = 28), and 19% for nonopioid optimization (n = 14). The most common primary diagnoses at consultation were for chronic low back pain (56%), chronic neck pain (20%), and osteoarthritis (16%).

The average MME at time of consult was 93 MME compared with 31 MME at discharge which was statisticially significant (P < .01) (Figure 1). The mean percent change in MME was 46%, including methadone and 42% excluding methadone. There was a 26% change in UDS, 28% change in informed consent, 85% change in PDMP, 194% change in naloxone, and 357% change in STORM reviews from baseline to discharge with all demonstrating statistical significance (P < .01) (Figure 2). At discharge, the most common opioid prescribed was morphine SA (short acting) (n = 10, 13%, 44 average MME) and oxycodone/acetaminophen (n = 10, 13%, 28 average MME) (Table 3).

The average number of days from consult to initial visit was 23 days (Table 4). Face-to-face was the primary means of initial visit with 92% (n = 69) of visits, but phone was the primary mode of follow-up with 73% of visits (n = 55). The average number of follow-up visits was 7, representing 176 average days of time in the Pharmacy Pain Clinic. Consultation to the behavioral health performance program was the most common referral (n = 13, 17%).

Discussion

There was a statistically significant decrease of the mean MME between the time of consult and the time of discharge. There also were statistically significant changes in use of opioid risk mitigation strategies. Since methadone has a high MME, the mean reduction of MME was calculated with methadone (46%) and without methadone (42%). These data are consistent with other published studies examining opioid tapers in the VA population. Harden and colleagues calculated a 46% mean reduction in MME over 12 months for 72 veterans from opioid tapers implemented by PCPs, pain service, or pharmacist-run clinics.¹⁹

There is controversy about equianalgesic doses and no established universal equianalgesic conversion calculator or dose. Numerous equianalgesic opioid dose calculators are available, but for this analysis the CDC MME conversion factors were used (available at: https://www.cdc.gov/drugoverdose/pdf/calculating_total_daily_dose-a.pdf). Previous literature compared existing calculators and found significant variances in calculated doses for methadone and fentanyl conversions.²⁰ Additionally, there have been concerns expressed with the safety of the CDC opioid calculator specifically surrounding the conversions for methadone and tapentadol.²¹ In the end, I chose the CDC calculator because it is established, readily available, and consistent.

Pharmacists in pain management can address access issues.^2,3,11,12 The average length of time from consult to initial visit was 23 days. Often patients may have seen a HCP who implemented a change at the time of consult and wanted the patient to be seen 1 month later. Many patients at the HSTMVH live far from the facility, making in-person visits difficult. A majority of the follow-up visits were conducted by telephone. Patients were offered all modalities available for follow-up, including telephone, in-person, or telemedicine, but patients most often picked telephone. Patients averaged 7 follow-up visits before discharge. This number of visits would have taken time from other health care team members who could have been addressing other veterans. Patients were seen in clinic for 176 days on average, which supports and follows recommendations for a slow, incremental taper.

The opioid medications prescribed changed over time in the clinic. Methadone prescriptions dropped from 20 to 6 at consult to discharge, and fentanyl prescriptions fell from 7 to 2, respectively. The CDC guideline suggests use of long-acting products with more predictable pharmacokinetics (eg, morphine SA or oxycodone SA) rather than fentanyl or methadone.¹⁵ Notably, the use of buprenorphine products with FDA approval for pain indications increased from consult to discharge. Many of the patients in this study had pulmonary comorbidities, placing them at higher risk for adverse outcomes. Buprenorphine is a partial μ-opioid receptor agonist with a ceiling on respiratory depression so is potentially less risky in those with pulmonary comorbidities.

The biggest changes in opioid risk mitigation occurred in PDMP, OEND program, and STORM reviews. An 85% increase in PDMP reviews occurred with referral to the clinic. Missouri is the only state without a state-run PDMP. However, the St. Louis County PDMP was developed based on city or county participation and encompasses 85% of the population of Missouri and 94% of HCPs in Missouri as of August 29, 2019.²² Because there is no state-level PDMP, a review of the St. Louis County PDMP was not required during the review period. Nevertheless, the Pharmacy Pain Clinic uses the St. Louis County PDMP at the initial visit and regularly during care. VA policy requires a specific note title be used to document each check of the PDMP.²³

There was a 194% increase in patients receiving naloxone with consultation to the Pharmacy Pain Clinic. Due to low coprescribing of naloxone for patients prescribed chronic opioid therapy, The author led an interdisciplinary team analysis of health care failure mode effects during the study period. This led to a process change with coprescribing of naloxone at refill in the primary care clinic.

The Comprehensive Addiction and Recovery Act of 2016 mandated that the VA review STORM on new start of opioids or patient identified as “very high-opioid prescription risk” category by an interdisciplinary opioid risk review team.²⁴ Thus many of the patients referred to clinic didn’t require STORM reviews since they were not new opioid starts or identified as high risk. However, in the standard review of all new patients to the Pharmacy Pain Clinic, a STORM review is conducted and documented to assess the patient’s level of risk.

Only 5 patients were started on opioid medications during the study period. This is consistent with both CDC and the joint VA/US Department of Defense opioid prescribing guidelines that recommend against initiation of opioids for chronic nonmalignant pain.^13,15 Two of the patients were prescribed tramadol for ineffective pain control at time of consult. Furthermore, 4 of the 5 patients were started on a short-acting opioid, which was supported by guidelines.^13,15 One patient was initiated on buprenorphine patches due to comorbid chronic kidney disease. The VA does not limit the quantity of new opioid prescriptions, although some states and private insurance plans are implementing limitations. Guidelines also recommend against exceeding 90 MME due to risk. The average MME in this project at discharge was 25 MME. Use of opioid risk mitigation for the new opioid starts was reasonable. The reason for the missing PDMP report is unknown based on chart review and atypical according to clinic practice.

Recently, efforts to expand pharmacist training and positions in pain management at VA facilities have been undertaken. In 2016, there were just 11 American Society of Health-System Pharmacists-accredited pharmacy postgraduate year 2 pain and palliative care residency programs, which has expanded to 26 sites in 2020.^2,3,25 In addition, the Clinical Pharmacy Practice Office and the VA Office of Rural Health have helped to hire 33 new pain management pharmacists.³

The role of pharmacists in prescribing controlled substances is limited mainly due to the small number of states that extend this authority.⁷ At the VA, a pharmacist can practice using any state of licensure. Therefore, a pharmacist working at a VA in a state that does not authorize controlled substance prescribing could obtain a license in a state that does permit it. However, the main barrier to obtaining other state licensures is the cost. At the time the author obtained controlled substance prescriptive authority, little direction was available on the process for advanced practice pharmacists at the VA. Since then, guidance has been developed to ease this process. Educational endeavors at VA have been implemented with the intent to increase the number of pharmacists with controlled substance prescriptive authority.

Barriers to pharmacists providing pain care extend beyond limited controlled substance prescriptive authority. Often pharmacists are still viewed in their traditional and operational role.^9,10 Other health care team members and patients may not be aware or familiar with the training, knowledge, and skills of pharmacist's and their suitability as an APP.^26,27 Most states permit pharmacists in establishing CDTA but not all. Additionally, some states recognize pharmacists as HCPs but many more do not. Furthermore, the Social Security Act does not include pharmacists as HCPs. This makes it challenging, though not impossible, for pharmacists to bill for their services.³

Strengths and Limitations

There were numerous strengths of the project. First, this addressed an unmet need in the literature with limited data discussing pharmacist prescribing controlled substances for pain management. There was 1 data reviewer who made the data collection process consistent. Since this retrospectively reviewed controlled substance prescribing in clinic, it captured real-world practice compared with that of experimental models. There were also several limitations in the project. The person collecting the data was also the person who conducted the clinic. The study was conducted retrospectively and based on documented information in the medical record. The population reviewed was primarily male and older, which fits the VA patient population but has less generalizability to other patient populations. This project was conducted at a single VA facility so may not be generalizable to other VA sites. It is unknown whether patients were again prescribed opioids if they left the VA for the community or another VA facility. The pain diagnoses or locations of pain were categorized to main groups and reliant on the referring provider. Another major weakness was the lack of comparison of pain scores or validated objective measure of function at baseline and at discharge. This consideration would be important for future work.

Conclusions

Pharmacists functioning as APP are key members of the pain management team. A review of a pharmacy-run pain clinic demonstrated statistically significant reduction in MME and improvement in opioid risk mitigation from consult to discharge. Patients enrolled in the pharmacy-managed clinic also had improvements in adherence to opioid risk mitigation strategies. Future attention should be focused on further expanding training and positions for pharmacists as APP in pain management.

Acknowledgments

The author thanks Chris Sedgwick for his assistance with data capture.

In the midst of an opioid overdose public health crisis, the US Department of Health and Human Services developed a 5-point strategy to combat this problem. One aspect of this strategy is improved pain management.¹ There is high demand for pain management services with a limited number of health care professionals appropriately trained to deliver care.² Pharmacists are integral members of the interdisciplinary pain team and meet this demand.

Background

For almost 50 years, pharmacists at the US Department of Veterans Affairs (VA) have been functioning as advanced practice providers (APP).³ Clinical pharmacy specialists (CPS) provide comprehensive medication management (CMM) and have a scope of practice (SOP). The SOP serves as the collaborating agreement and outlines the clinical duties permitted in delivering patient care. In addition, the SOP may indicate specific practice areas and are standardized across VA (Table 1).^4,5 Pharmacists apply for a SOP and must prove their competency in the practice area and provide documentation of their education, training, experience, knowledge, and skills.^5,6 Residency and/or board certification are not required though helpful. A pharmacist’s SOP is reviewed and approved by the facility executive committee.⁵ Pharmacists with a SOP undergo professional practice evaluation twice a year. Prescribing controlled substances is permissible in the SOP if approved by the facility and allowed by the state of licensure. According to the US Drug Enforcement Agency (DEA) as of February 10, 2020, 8 states (California, Washington, Idaho, Massachusetts, Montana, New Mexico, North Carolina, and Ohio) allow pharmacists to prescribe controlled substances.⁷

The VA developed the Pharmacists Achieve Results with Medications Documentation (PhARMD) tool that allows clinical pharmacists to document specific interventions made during clinical care and is included in their progress note. Data from fiscal year 2017 demonstrates that 136,041 pain management interventions were made by pharmacists across VA. The majority of these interventions were implemented by a CPS working autonomously as an APP.⁸

Several articles discuss the pharmacists role in the opioid crisis, although no outcomes data were provided. Chisholm-Burns and colleagues listed multiple potential ways that pharmacists can intervene, including managing pain in primary care clinic settings by using collaborative drug therapy agreements (CDTAs), using opioid exit plans and discharge planning in collaboration with other health care providers (HCPs), or making recommendations to the prescribers before writing prescriptions.⁹ Compton and colleagues similarly reviewed pharmacist roles in the opioid crisis. However, their focus was on dispensing pharmacists that provided education to patients about storage and disposal of opioids, identified opioid misuse, provided opioid overdose education and naloxone, and checked prescription drug monitoring programs (PDMPs).¹⁰ Missing from these articles was the role of the clinical pharmacist working as an APP delivering direct patient care and prescribing controlled substances.

Hammer and colleagues discussed the role of an oncology CPS with controlled substance prescriptive authority in pain management at an outpatient cancer center in Washington state.¹¹ Under a CDTA, pharmacists could prescribe medications, including controlled substances if they obtain DEA registration. The pharmacist completed a comprehensive in-person assessment. The attending physician conducted a physical examination. Then the pharmacist presented the patient and proposed regimen to the interprofessional team to determine a final plan. Ultimately, the pharmacist wrote any controlled substance prescriptions. The patient followed up every 1 to 4 weeks by telephone with a nurse, and in-person assessments occurred at least every 6 months. No outcomes data were provided.¹¹

Dole and colleagues reviewed the role of a pharmacist who had controlled substance prescriptive authority in a pain management clinic. The pharmacist provider saw up to 18 patients a day and then managed refill requests for 3 hours a day. The main outcome was change in visual analog scale (VAS) pain scores. Findings showed that reductions in VAS pain scores were statistically significant (P < .01). The pharmacist processed about 150 refills with an unclear number of controlled substances requests a day based on a medication-refill protocol. This was felt to improve access to physicians for acute needs, improve consistency in refills, and capture patients in need of follow-up. Additionally, the clinic saved $455,238 after 1 year.¹²

Study Aims

A review of the literature indicated sparse data on the impact of a pharmacist on opioid tapering, opioid dose, and opioid risk mitigation when the pharmacist is prescribing controlled substances. The purpose of this retrospective review was to characterize the controlled substance prescribing practices by the pharmacy pain clinic. The aim was to examine the pharmacist impact on morphine milligram equivalent (MME) and compliance with opioid risk mitigation strategies.

Methods

This project was a retrospective, single-center, chart review. The project was reviewed and approved by the University of Missouri-Columbia Institutional Review Board used by the Harry S. Truman Memorial Veterans’ Hospital (HSTMVH) as a quality improvement project. The author applied for controlled substance registration through the DEA and was issued registration April 30, 2018. The State of Ohio Board of Pharmacy was contacted as required by Ohio Administrative Code. The author's updated SOP to allow controlled substance prescribing was approved July 23, 2018. The CPS functions as an APP within an interdisciplinary pain management team that includes physicians, occupational and physical therapists, complementary and integrative health, and a psychologist. The reason for Pharmacy Pain Consult is required and it is primarily submitted through the electronic health record. The consult is reviewed for appropriateness and once approved is scheduled by support staff. Once the patient is stabilized, the patient is discharged back to their primary care provider (PCP) or referring provider for continued care. Patients were considered stabilized when their patient-specific goals were met, which varied from use of the lowest effective opioid dose to taper to discontinuation of opioids with no further medication changes needed. The taper strategy for each patient was individualized. Patients were generally tapered on their existing opioid medication unless they were new to the VA and on nonformulary medications or experiencing a significant adverse reaction. Numerous references are available through VA to assist with opioid tapering.^13,14 The CPS is able to refer patients to other services, including behavioral health for substance use disorder treatment and medication-assisted treatment if concerns were identified.

Initial data were collected from the Veterans Integrated Service Network (VISN) 15 Corporate Data Warehouse by the VISN pharmacy analytics program manager. The original report included patients prescribed a Schedule II to V controlled substance by the author from July 1, 2018 to January 31, 2020. Chart review was conducted on each patient to obtain additional data. At the time of consult and discharge the following data were collected: opioid medication; MME; use of opioid risk mitigation strategies, such as urine drug screens (UDS), informed consent, opioid overdose education and naloxone distribution program (OEND), risk assessment via stratification tool for opioid risk mitigation (STORM), PDMP checks; and nonopioid medication number and classes.

Patients were included in the review if they were prescribed an opioid Schedule II or III controlled substance between July 1, 2018 and January 31, 2020. Patient were excluded if they were prescribed an opioid Schedule II or III controlled substance primarily as coverage for another prescriber. Patients prescribed only pregabalin, tramadol, or a benzodiazepine also were excluded.

The primary endpoint was change in MME from baseline to discharge from clinic. Secondary endpoints included change in opioid risk mitigation strategies and change in opioid medications prescribed from baseline to discharge.

Descriptive statistics were used to analyze parts of the data. A 2-sided t test was used to compare baseline and discharge MME. The Fisher exact test was used to compare nominal data of opioid risk mitigation strategies.

Calculation of MME was performed using the conversion factors provided by the Centers Disease Control and Prevention (CDC) for opioid guideline.¹⁵ For buprenorphine, tapentadol, and levorphanol conversion ratios were obtained from other sources. The conversion ratios used, included 75:1 for oral morphine to transdermal buprenorphine, 1:3.3 for oral morphine to oral tapentadol, and 1:7.5 for oral levorphanol to oral morphine.^16,17 The Revised Standards for Quality Improvement Reporting Excellence (SQUIRE 2.0) was used to write the manuscript.¹⁸

Results

Seventy-five patients were included in this review. The average age of patients was 66 years; and 12% were female (n = 9) (Table 2). The largest number of consults came from PCPs (44%, n = 33) and the pain clinic (43%, n = 32). Nearly half (48%) of the consultations were for opioid tapering (n = 36), followed by 37% for opioid optimization or monitoring (n = 28), and 19% for nonopioid optimization (n = 14). The most common primary diagnoses at consultation were for chronic low back pain (56%), chronic neck pain (20%), and osteoarthritis (16%).

The average MME at time of consult was 93 MME compared with 31 MME at discharge which was statisticially significant (P < .01) (Figure 1). The mean percent change in MME was 46%, including methadone and 42% excluding methadone. There was a 26% change in UDS, 28% change in informed consent, 85% change in PDMP, 194% change in naloxone, and 357% change in STORM reviews from baseline to discharge with all demonstrating statistical significance (P < .01) (Figure 2). At discharge, the most common opioid prescribed was morphine SA (short acting) (n = 10, 13%, 44 average MME) and oxycodone/acetaminophen (n = 10, 13%, 28 average MME) (Table 3).

The average number of days from consult to initial visit was 23 days (Table 4). Face-to-face was the primary means of initial visit with 92% (n = 69) of visits, but phone was the primary mode of follow-up with 73% of visits (n = 55). The average number of follow-up visits was 7, representing 176 average days of time in the Pharmacy Pain Clinic. Consultation to the behavioral health performance program was the most common referral (n = 13, 17%).

Discussion

There was a statistically significant decrease of the mean MME between the time of consult and the time of discharge. There also were statistically significant changes in use of opioid risk mitigation strategies. Since methadone has a high MME, the mean reduction of MME was calculated with methadone (46%) and without methadone (42%). These data are consistent with other published studies examining opioid tapers in the VA population. Harden and colleagues calculated a 46% mean reduction in MME over 12 months for 72 veterans from opioid tapers implemented by PCPs, pain service, or pharmacist-run clinics.¹⁹

There is controversy about equianalgesic doses and no established universal equianalgesic conversion calculator or dose. Numerous equianalgesic opioid dose calculators are available, but for this analysis the CDC MME conversion factors were used (available at: https://www.cdc.gov/drugoverdose/pdf/calculating_total_daily_dose-a.pdf). Previous literature compared existing calculators and found significant variances in calculated doses for methadone and fentanyl conversions.²⁰ Additionally, there have been concerns expressed with the safety of the CDC opioid calculator specifically surrounding the conversions for methadone and tapentadol.²¹ In the end, I chose the CDC calculator because it is established, readily available, and consistent.

Pharmacists in pain management can address access issues.^2,3,11,12 The average length of time from consult to initial visit was 23 days. Often patients may have seen a HCP who implemented a change at the time of consult and wanted the patient to be seen 1 month later. Many patients at the HSTMVH live far from the facility, making in-person visits difficult. A majority of the follow-up visits were conducted by telephone. Patients were offered all modalities available for follow-up, including telephone, in-person, or telemedicine, but patients most often picked telephone. Patients averaged 7 follow-up visits before discharge. This number of visits would have taken time from other health care team members who could have been addressing other veterans. Patients were seen in clinic for 176 days on average, which supports and follows recommendations for a slow, incremental taper.

The opioid medications prescribed changed over time in the clinic. Methadone prescriptions dropped from 20 to 6 at consult to discharge, and fentanyl prescriptions fell from 7 to 2, respectively. The CDC guideline suggests use of long-acting products with more predictable pharmacokinetics (eg, morphine SA or oxycodone SA) rather than fentanyl or methadone.¹⁵ Notably, the use of buprenorphine products with FDA approval for pain indications increased from consult to discharge. Many of the patients in this study had pulmonary comorbidities, placing them at higher risk for adverse outcomes. Buprenorphine is a partial μ-opioid receptor agonist with a ceiling on respiratory depression so is potentially less risky in those with pulmonary comorbidities.

The biggest changes in opioid risk mitigation occurred in PDMP, OEND program, and STORM reviews. An 85% increase in PDMP reviews occurred with referral to the clinic. Missouri is the only state without a state-run PDMP. However, the St. Louis County PDMP was developed based on city or county participation and encompasses 85% of the population of Missouri and 94% of HCPs in Missouri as of August 29, 2019.²² Because there is no state-level PDMP, a review of the St. Louis County PDMP was not required during the review period. Nevertheless, the Pharmacy Pain Clinic uses the St. Louis County PDMP at the initial visit and regularly during care. VA policy requires a specific note title be used to document each check of the PDMP.²³

There was a 194% increase in patients receiving naloxone with consultation to the Pharmacy Pain Clinic. Due to low coprescribing of naloxone for patients prescribed chronic opioid therapy, The author led an interdisciplinary team analysis of health care failure mode effects during the study period. This led to a process change with coprescribing of naloxone at refill in the primary care clinic.

The Comprehensive Addiction and Recovery Act of 2016 mandated that the VA review STORM on new start of opioids or patient identified as “very high-opioid prescription risk” category by an interdisciplinary opioid risk review team.²⁴ Thus many of the patients referred to clinic didn’t require STORM reviews since they were not new opioid starts or identified as high risk. However, in the standard review of all new patients to the Pharmacy Pain Clinic, a STORM review is conducted and documented to assess the patient’s level of risk.

Only 5 patients were started on opioid medications during the study period. This is consistent with both CDC and the joint VA/US Department of Defense opioid prescribing guidelines that recommend against initiation of opioids for chronic nonmalignant pain.^13,15 Two of the patients were prescribed tramadol for ineffective pain control at time of consult. Furthermore, 4 of the 5 patients were started on a short-acting opioid, which was supported by guidelines.^13,15 One patient was initiated on buprenorphine patches due to comorbid chronic kidney disease. The VA does not limit the quantity of new opioid prescriptions, although some states and private insurance plans are implementing limitations. Guidelines also recommend against exceeding 90 MME due to risk. The average MME in this project at discharge was 25 MME. Use of opioid risk mitigation for the new opioid starts was reasonable. The reason for the missing PDMP report is unknown based on chart review and atypical according to clinic practice.

Recently, efforts to expand pharmacist training and positions in pain management at VA facilities have been undertaken. In 2016, there were just 11 American Society of Health-System Pharmacists-accredited pharmacy postgraduate year 2 pain and palliative care residency programs, which has expanded to 26 sites in 2020.^2,3,25 In addition, the Clinical Pharmacy Practice Office and the VA Office of Rural Health have helped to hire 33 new pain management pharmacists.³

The role of pharmacists in prescribing controlled substances is limited mainly due to the small number of states that extend this authority.⁷ At the VA, a pharmacist can practice using any state of licensure. Therefore, a pharmacist working at a VA in a state that does not authorize controlled substance prescribing could obtain a license in a state that does permit it. However, the main barrier to obtaining other state licensures is the cost. At the time the author obtained controlled substance prescriptive authority, little direction was available on the process for advanced practice pharmacists at the VA. Since then, guidance has been developed to ease this process. Educational endeavors at VA have been implemented with the intent to increase the number of pharmacists with controlled substance prescriptive authority.

Barriers to pharmacists providing pain care extend beyond limited controlled substance prescriptive authority. Often pharmacists are still viewed in their traditional and operational role.^9,10 Other health care team members and patients may not be aware or familiar with the training, knowledge, and skills of pharmacist's and their suitability as an APP.^26,27 Most states permit pharmacists in establishing CDTA but not all. Additionally, some states recognize pharmacists as HCPs but many more do not. Furthermore, the Social Security Act does not include pharmacists as HCPs. This makes it challenging, though not impossible, for pharmacists to bill for their services.³

Strengths and Limitations

There were numerous strengths of the project. First, this addressed an unmet need in the literature with limited data discussing pharmacist prescribing controlled substances for pain management. There was 1 data reviewer who made the data collection process consistent. Since this retrospectively reviewed controlled substance prescribing in clinic, it captured real-world practice compared with that of experimental models. There were also several limitations in the project. The person collecting the data was also the person who conducted the clinic. The study was conducted retrospectively and based on documented information in the medical record. The population reviewed was primarily male and older, which fits the VA patient population but has less generalizability to other patient populations. This project was conducted at a single VA facility so may not be generalizable to other VA sites. It is unknown whether patients were again prescribed opioids if they left the VA for the community or another VA facility. The pain diagnoses or locations of pain were categorized to main groups and reliant on the referring provider. Another major weakness was the lack of comparison of pain scores or validated objective measure of function at baseline and at discharge. This consideration would be important for future work.

Conclusions

Pharmacists functioning as APP are key members of the pain management team. A review of a pharmacy-run pain clinic demonstrated statistically significant reduction in MME and improvement in opioid risk mitigation from consult to discharge. Patients enrolled in the pharmacy-managed clinic also had improvements in adherence to opioid risk mitigation strategies. Future attention should be focused on further expanding training and positions for pharmacists as APP in pain management.

Acknowledgments

The author thanks Chris Sedgwick for his assistance with data capture.

In the midst of an opioid overdose public health crisis, the US Department of Health and Human Services developed a 5-point strategy to combat this problem. One aspect of this strategy is improved pain management.¹ There is high demand for pain management services with a limited number of health care professionals appropriately trained to deliver care.² Pharmacists are integral members of the interdisciplinary pain team and meet this demand.

Background

For almost 50 years, pharmacists at the US Department of Veterans Affairs (VA) have been functioning as advanced practice providers (APP).³ Clinical pharmacy specialists (CPS) provide comprehensive medication management (CMM) and have a scope of practice (SOP). The SOP serves as the collaborating agreement and outlines the clinical duties permitted in delivering patient care. In addition, the SOP may indicate specific practice areas and are standardized across VA (Table 1).^4,5 Pharmacists apply for a SOP and must prove their competency in the practice area and provide documentation of their education, training, experience, knowledge, and skills.^5,6 Residency and/or board certification are not required though helpful. A pharmacist’s SOP is reviewed and approved by the facility executive committee.⁵ Pharmacists with a SOP undergo professional practice evaluation twice a year. Prescribing controlled substances is permissible in the SOP if approved by the facility and allowed by the state of licensure. According to the US Drug Enforcement Agency (DEA) as of February 10, 2020, 8 states (California, Washington, Idaho, Massachusetts, Montana, New Mexico, North Carolina, and Ohio) allow pharmacists to prescribe controlled substances.⁷

The VA developed the Pharmacists Achieve Results with Medications Documentation (PhARMD) tool that allows clinical pharmacists to document specific interventions made during clinical care and is included in their progress note. Data from fiscal year 2017 demonstrates that 136,041 pain management interventions were made by pharmacists across VA. The majority of these interventions were implemented by a CPS working autonomously as an APP.⁸

Several articles discuss the pharmacists role in the opioid crisis, although no outcomes data were provided. Chisholm-Burns and colleagues listed multiple potential ways that pharmacists can intervene, including managing pain in primary care clinic settings by using collaborative drug therapy agreements (CDTAs), using opioid exit plans and discharge planning in collaboration with other health care providers (HCPs), or making recommendations to the prescribers before writing prescriptions.⁹ Compton and colleagues similarly reviewed pharmacist roles in the opioid crisis. However, their focus was on dispensing pharmacists that provided education to patients about storage and disposal of opioids, identified opioid misuse, provided opioid overdose education and naloxone, and checked prescription drug monitoring programs (PDMPs).¹⁰ Missing from these articles was the role of the clinical pharmacist working as an APP delivering direct patient care and prescribing controlled substances.

Hammer and colleagues discussed the role of an oncology CPS with controlled substance prescriptive authority in pain management at an outpatient cancer center in Washington state.¹¹ Under a CDTA, pharmacists could prescribe medications, including controlled substances if they obtain DEA registration. The pharmacist completed a comprehensive in-person assessment. The attending physician conducted a physical examination. Then the pharmacist presented the patient and proposed regimen to the interprofessional team to determine a final plan. Ultimately, the pharmacist wrote any controlled substance prescriptions. The patient followed up every 1 to 4 weeks by telephone with a nurse, and in-person assessments occurred at least every 6 months. No outcomes data were provided.¹¹

Dole and colleagues reviewed the role of a pharmacist who had controlled substance prescriptive authority in a pain management clinic. The pharmacist provider saw up to 18 patients a day and then managed refill requests for 3 hours a day. The main outcome was change in visual analog scale (VAS) pain scores. Findings showed that reductions in VAS pain scores were statistically significant (P < .01). The pharmacist processed about 150 refills with an unclear number of controlled substances requests a day based on a medication-refill protocol. This was felt to improve access to physicians for acute needs, improve consistency in refills, and capture patients in need of follow-up. Additionally, the clinic saved $455,238 after 1 year.¹²

Study Aims

A review of the literature indicated sparse data on the impact of a pharmacist on opioid tapering, opioid dose, and opioid risk mitigation when the pharmacist is prescribing controlled substances. The purpose of this retrospective review was to characterize the controlled substance prescribing practices by the pharmacy pain clinic. The aim was to examine the pharmacist impact on morphine milligram equivalent (MME) and compliance with opioid risk mitigation strategies.

Methods

This project was a retrospective, single-center, chart review. The project was reviewed and approved by the University of Missouri-Columbia Institutional Review Board used by the Harry S. Truman Memorial Veterans’ Hospital (HSTMVH) as a quality improvement project. The author applied for controlled substance registration through the DEA and was issued registration April 30, 2018. The State of Ohio Board of Pharmacy was contacted as required by Ohio Administrative Code. The author's updated SOP to allow controlled substance prescribing was approved July 23, 2018. The CPS functions as an APP within an interdisciplinary pain management team that includes physicians, occupational and physical therapists, complementary and integrative health, and a psychologist. The reason for Pharmacy Pain Consult is required and it is primarily submitted through the electronic health record. The consult is reviewed for appropriateness and once approved is scheduled by support staff. Once the patient is stabilized, the patient is discharged back to their primary care provider (PCP) or referring provider for continued care. Patients were considered stabilized when their patient-specific goals were met, which varied from use of the lowest effective opioid dose to taper to discontinuation of opioids with no further medication changes needed. The taper strategy for each patient was individualized. Patients were generally tapered on their existing opioid medication unless they were new to the VA and on nonformulary medications or experiencing a significant adverse reaction. Numerous references are available through VA to assist with opioid tapering.^13,14 The CPS is able to refer patients to other services, including behavioral health for substance use disorder treatment and medication-assisted treatment if concerns were identified.

Initial data were collected from the Veterans Integrated Service Network (VISN) 15 Corporate Data Warehouse by the VISN pharmacy analytics program manager. The original report included patients prescribed a Schedule II to V controlled substance by the author from July 1, 2018 to January 31, 2020. Chart review was conducted on each patient to obtain additional data. At the time of consult and discharge the following data were collected: opioid medication; MME; use of opioid risk mitigation strategies, such as urine drug screens (UDS), informed consent, opioid overdose education and naloxone distribution program (OEND), risk assessment via stratification tool for opioid risk mitigation (STORM), PDMP checks; and nonopioid medication number and classes.

Patients were included in the review if they were prescribed an opioid Schedule II or III controlled substance between July 1, 2018 and January 31, 2020. Patient were excluded if they were prescribed an opioid Schedule II or III controlled substance primarily as coverage for another prescriber. Patients prescribed only pregabalin, tramadol, or a benzodiazepine also were excluded.

The primary endpoint was change in MME from baseline to discharge from clinic. Secondary endpoints included change in opioid risk mitigation strategies and change in opioid medications prescribed from baseline to discharge.

Descriptive statistics were used to analyze parts of the data. A 2-sided t test was used to compare baseline and discharge MME. The Fisher exact test was used to compare nominal data of opioid risk mitigation strategies.

Calculation of MME was performed using the conversion factors provided by the Centers Disease Control and Prevention (CDC) for opioid guideline.¹⁵ For buprenorphine, tapentadol, and levorphanol conversion ratios were obtained from other sources. The conversion ratios used, included 75:1 for oral morphine to transdermal buprenorphine, 1:3.3 for oral morphine to oral tapentadol, and 1:7.5 for oral levorphanol to oral morphine.^16,17 The Revised Standards for Quality Improvement Reporting Excellence (SQUIRE 2.0) was used to write the manuscript.¹⁸

Results

Seventy-five patients were included in this review. The average age of patients was 66 years; and 12% were female (n = 9) (Table 2). The largest number of consults came from PCPs (44%, n = 33) and the pain clinic (43%, n = 32). Nearly half (48%) of the consultations were for opioid tapering (n = 36), followed by 37% for opioid optimization or monitoring (n = 28), and 19% for nonopioid optimization (n = 14). The most common primary diagnoses at consultation were for chronic low back pain (56%), chronic neck pain (20%), and osteoarthritis (16%).

The average MME at time of consult was 93 MME compared with 31 MME at discharge which was statisticially significant (P < .01) (Figure 1). The mean percent change in MME was 46%, including methadone and 42% excluding methadone. There was a 26% change in UDS, 28% change in informed consent, 85% change in PDMP, 194% change in naloxone, and 357% change in STORM reviews from baseline to discharge with all demonstrating statistical significance (P < .01) (Figure 2). At discharge, the most common opioid prescribed was morphine SA (short acting) (n = 10, 13%, 44 average MME) and oxycodone/acetaminophen (n = 10, 13%, 28 average MME) (Table 3).

The average number of days from consult to initial visit was 23 days (Table 4). Face-to-face was the primary means of initial visit with 92% (n = 69) of visits, but phone was the primary mode of follow-up with 73% of visits (n = 55). The average number of follow-up visits was 7, representing 176 average days of time in the Pharmacy Pain Clinic. Consultation to the behavioral health performance program was the most common referral (n = 13, 17%).

Discussion

There was a statistically significant decrease of the mean MME between the time of consult and the time of discharge. There also were statistically significant changes in use of opioid risk mitigation strategies. Since methadone has a high MME, the mean reduction of MME was calculated with methadone (46%) and without methadone (42%). These data are consistent with other published studies examining opioid tapers in the VA population. Harden and colleagues calculated a 46% mean reduction in MME over 12 months for 72 veterans from opioid tapers implemented by PCPs, pain service, or pharmacist-run clinics.¹⁹

There is controversy about equianalgesic doses and no established universal equianalgesic conversion calculator or dose. Numerous equianalgesic opioid dose calculators are available, but for this analysis the CDC MME conversion factors were used (available at: https://www.cdc.gov/drugoverdose/pdf/calculating_total_daily_dose-a.pdf). Previous literature compared existing calculators and found significant variances in calculated doses for methadone and fentanyl conversions.²⁰ Additionally, there have been concerns expressed with the safety of the CDC opioid calculator specifically surrounding the conversions for methadone and tapentadol.²¹ In the end, I chose the CDC calculator because it is established, readily available, and consistent.

Pharmacists in pain management can address access issues.^2,3,11,12 The average length of time from consult to initial visit was 23 days. Often patients may have seen a HCP who implemented a change at the time of consult and wanted the patient to be seen 1 month later. Many patients at the HSTMVH live far from the facility, making in-person visits difficult. A majority of the follow-up visits were conducted by telephone. Patients were offered all modalities available for follow-up, including telephone, in-person, or telemedicine, but patients most often picked telephone. Patients averaged 7 follow-up visits before discharge. This number of visits would have taken time from other health care team members who could have been addressing other veterans. Patients were seen in clinic for 176 days on average, which supports and follows recommendations for a slow, incremental taper.

The opioid medications prescribed changed over time in the clinic. Methadone prescriptions dropped from 20 to 6 at consult to discharge, and fentanyl prescriptions fell from 7 to 2, respectively. The CDC guideline suggests use of long-acting products with more predictable pharmacokinetics (eg, morphine SA or oxycodone SA) rather than fentanyl or methadone.¹⁵ Notably, the use of buprenorphine products with FDA approval for pain indications increased from consult to discharge. Many of the patients in this study had pulmonary comorbidities, placing them at higher risk for adverse outcomes. Buprenorphine is a partial μ-opioid receptor agonist with a ceiling on respiratory depression so is potentially less risky in those with pulmonary comorbidities.

The biggest changes in opioid risk mitigation occurred in PDMP, OEND program, and STORM reviews. An 85% increase in PDMP reviews occurred with referral to the clinic. Missouri is the only state without a state-run PDMP. However, the St. Louis County PDMP was developed based on city or county participation and encompasses 85% of the population of Missouri and 94% of HCPs in Missouri as of August 29, 2019.²² Because there is no state-level PDMP, a review of the St. Louis County PDMP was not required during the review period. Nevertheless, the Pharmacy Pain Clinic uses the St. Louis County PDMP at the initial visit and regularly during care. VA policy requires a specific note title be used to document each check of the PDMP.²³

There was a 194% increase in patients receiving naloxone with consultation to the Pharmacy Pain Clinic. Due to low coprescribing of naloxone for patients prescribed chronic opioid therapy, The author led an interdisciplinary team analysis of health care failure mode effects during the study period. This led to a process change with coprescribing of naloxone at refill in the primary care clinic.

The Comprehensive Addiction and Recovery Act of 2016 mandated that the VA review STORM on new start of opioids or patient identified as “very high-opioid prescription risk” category by an interdisciplinary opioid risk review team.²⁴ Thus many of the patients referred to clinic didn’t require STORM reviews since they were not new opioid starts or identified as high risk. However, in the standard review of all new patients to the Pharmacy Pain Clinic, a STORM review is conducted and documented to assess the patient’s level of risk.

Only 5 patients were started on opioid medications during the study period. This is consistent with both CDC and the joint VA/US Department of Defense opioid prescribing guidelines that recommend against initiation of opioids for chronic nonmalignant pain.^13,15 Two of the patients were prescribed tramadol for ineffective pain control at time of consult. Furthermore, 4 of the 5 patients were started on a short-acting opioid, which was supported by guidelines.^13,15 One patient was initiated on buprenorphine patches due to comorbid chronic kidney disease. The VA does not limit the quantity of new opioid prescriptions, although some states and private insurance plans are implementing limitations. Guidelines also recommend against exceeding 90 MME due to risk. The average MME in this project at discharge was 25 MME. Use of opioid risk mitigation for the new opioid starts was reasonable. The reason for the missing PDMP report is unknown based on chart review and atypical according to clinic practice.

Recently, efforts to expand pharmacist training and positions in pain management at VA facilities have been undertaken. In 2016, there were just 11 American Society of Health-System Pharmacists-accredited pharmacy postgraduate year 2 pain and palliative care residency programs, which has expanded to 26 sites in 2020.^2,3,25 In addition, the Clinical Pharmacy Practice Office and the VA Office of Rural Health have helped to hire 33 new pain management pharmacists.³

The role of pharmacists in prescribing controlled substances is limited mainly due to the small number of states that extend this authority.⁷ At the VA, a pharmacist can practice using any state of licensure. Therefore, a pharmacist working at a VA in a state that does not authorize controlled substance prescribing could obtain a license in a state that does permit it. However, the main barrier to obtaining other state licensures is the cost. At the time the author obtained controlled substance prescriptive authority, little direction was available on the process for advanced practice pharmacists at the VA. Since then, guidance has been developed to ease this process. Educational endeavors at VA have been implemented with the intent to increase the number of pharmacists with controlled substance prescriptive authority.

Barriers to pharmacists providing pain care extend beyond limited controlled substance prescriptive authority. Often pharmacists are still viewed in their traditional and operational role.^9,10 Other health care team members and patients may not be aware or familiar with the training, knowledge, and skills of pharmacist's and their suitability as an APP.^26,27 Most states permit pharmacists in establishing CDTA but not all. Additionally, some states recognize pharmacists as HCPs but many more do not. Furthermore, the Social Security Act does not include pharmacists as HCPs. This makes it challenging, though not impossible, for pharmacists to bill for their services.³

Strengths and Limitations

There were numerous strengths of the project. First, this addressed an unmet need in the literature with limited data discussing pharmacist prescribing controlled substances for pain management. There was 1 data reviewer who made the data collection process consistent. Since this retrospectively reviewed controlled substance prescribing in clinic, it captured real-world practice compared with that of experimental models. There were also several limitations in the project. The person collecting the data was also the person who conducted the clinic. The study was conducted retrospectively and based on documented information in the medical record. The population reviewed was primarily male and older, which fits the VA patient population but has less generalizability to other patient populations. This project was conducted at a single VA facility so may not be generalizable to other VA sites. It is unknown whether patients were again prescribed opioids if they left the VA for the community or another VA facility. The pain diagnoses or locations of pain were categorized to main groups and reliant on the referring provider. Another major weakness was the lack of comparison of pain scores or validated objective measure of function at baseline and at discharge. This consideration would be important for future work.

Conclusions

Pharmacists functioning as APP are key members of the pain management team. A review of a pharmacy-run pain clinic demonstrated statistically significant reduction in MME and improvement in opioid risk mitigation from consult to discharge. Patients enrolled in the pharmacy-managed clinic also had improvements in adherence to opioid risk mitigation strategies. Future attention should be focused on further expanding training and positions for pharmacists as APP in pain management.

Acknowledgments

The author thanks Chris Sedgwick for his assistance with data capture.